Book: Profiting from Hedge Funds: Winning Strategies for the Little Guy
Author: John Konnayil Vincent
Difficulty: Intermediate
The Big Idea
A public filing can give you an investment idea. It cannot give you the manager’s entire portfolio, execution history or reasons for owning it. The useful skill is turning that incomplete evidence into a dated, testable research process, then checking the business and the risk yourself.
Vincent’s book examines selected managers, develops mechanical ways to follow reported holdings, and introduces fundamental analysis and portfolio sizing. This independent companion covers all 22 substantive chapters with original lessons, examples, exercises and explained answers. Keep the book beside you: reading assignments and source-evidence tasks are essential to the workbook. This is not an abridgment or a replacement for the original.
Contents
- Quick revision sheet
- Reading instructions and evidence rules
- Current filing rules and important corrections
- Seven flexible study blocks
- Complete chapter map
- Part I: tracking the managers, Chapters 1–13
- Part II: mechanical models, Chapters 14–18
- Part III: independent analysis, Chapters 19–22
- Chapter 20: additional financial-analysis labs
- Compare the models before testing
- Pine indicator and manual observation lab
- Four reusable worksheets
- Capstone and grading rubric
- Technical and current-context references
- Final Takeaway
Quick Revision Sheet
| Concept | Use it this way | Avoid this inference |
|---|---|---|
| 13F snapshot | Record the reporting date and actual public timestamp separately. | A quarter-end holding proves the position still exists today. |
| Net share addition | Adjusted closing shares minus comparable prior shares; value consistently if ranking. | A rise in market value necessarily means buying. |
| Issuer ownership | Shares beneficially owned / relevant issuer shares outstanding. | Owning 5% of an issuer means a 5% portfolio allocation. |
| Sleeve weight | Position value / explicitly defined eligible sleeve value. | A large sleeve position is equally large relative to all assets. |
| Equal weighting | Each of N holdings receives 1/N of the allocated sleeve. | Three famous names provide adequate diversification. |
| Score weighting | Target weight = score / sum of included scores. | A score of 3 is an allocation of 3%. |
| Retained-weight normalization | Retained source weight / sum of retained source weights. | Dropping small positions leaves the other target percentages unchanged. |
| Margin of safety | (Estimated value − price) / estimated value; identify the estimate and scenario. | A valuation discount is a guaranteed maximum loss. |
| Diversification | Check shared drivers, covariance, liquidity and concentration. | Zero correlation proves independence or correlations cannot change. |
| Research result | Evaluate frozen rules on later, untouched data after realistic costs. | A selected historical winner establishes a future edge. |
| Measure | Formula | Qualification |
|---|---|---|
| Simple return | End value / start value − 1 | Adjust for distributions and external cash flows appropriately. |
| CAGR | (End wealth / start wealth)1/years − 1 | Requires a comparable no-external-flow wealth path. Count elapsed intervals correctly. |
| Whole-portfolio contribution | Initial weight × holding return | Single-period approximation with no intervening rebalancing. |
| Gordon equity value | Next dividend / (required equity return − perpetual growth) | Use compatible annual units, with required return > growth. |
| Two-asset volatility | √[w₁²σ₁² + w₂²σ₂² + 2w₁w₂ρσ₁σ₂] | A historical covariance estimate is not a tail-loss guarantee. |
| Drawdown | 1 − current wealth / prior-or-current running peak | Use an ordered wealth path, not only the final return. |
| One-way turnover convention | ½ × sum of absolute target-minus-pretrade weight changes | Include cash in the weights; record this convention and actual dollar orders. |
Reading Instructions and Evidence Rules
The supplied source is the 2013 Wiley book, cloth ISBN 9781118465202. Assignments below use its printed page numbers; in this PDF, Chapter 1 starts on viewer page 31, corresponding to printed page 3. Part dividers, references and blanks account for gaps. The preface is useful orientation; the index and author biography are reference material rather than extra lesson chapters.
Read the assigned chapter first. Then work its example, answer the exercise with the key closed and return to the book to record one specific supporting passage or table and one limitation. The chapter map supplies direct links. Basic percentages, spreadsheet formulas and familiarity with financial statements are enough to begin; the valuation labs may take several sittings.
Manager profiles, organizational titles, holdings and test periods belong to the book’s historical setting, mostly ending around 2012. They are not current holdings, current job descriptions or current stock recommendations.
All fictional numbers, exercises, worksheets, capstone and chart implementation here are ours. They are not the author’s reported results, proprietary manager methods or evidence of profitable live execution.
For each return claim in the book, record the selection date, start/end dates, dividends, benchmark type and trading assumptions. Its selected-manager cases and short historical models are useful to critique; they do not justify language such as “guaranteed,” “foolproof” or “highly likely” without independent evidence.
Current Filing Rules and Important Corrections
Current-context checks below were made against SEC guidance on November 2, 2026. These are reading aids, not a filing-compliance opinion. Follow the linked current instructions for a live case.
| Topic | Reading correction | Primary reference |
|---|---|---|
| 13F timing and units | Quarterly deadline: within 45 days, subject to calendar rules; not a mandatory 45-day minimum delay. Since January 3, 2023, reported values round to dollars, replacing thousands. | SEC 13F FAQ |
| 13F coverage | Applies to qualifying managers and Section 13(f) securities, not every manager or asset. Reconcile shared reporting. Some eligible securities are not ordinary U.S. shares. | SEC Form 13F instructions |
| Options and missing exposure | Short equity and written options are excluded. Reported long-option rows do not provide strikes or expiries; underlying values are not option premiums. A filing is not a full exposure map. | SEC 13F FAQ |
| Schedule 13D | Initial disclosure generally within five business days after the triggering acquisition; material-change amendments within two business days. The book’s ten-day rule is historical. | SEC beneficial-ownership interpretations |
| Schedule 13G | Eligibility and deadlines depend on filer category and thresholds. Do not apply one blanket 13D/13G deadline. | SEC beneficial-ownership interpretations |
| Forms 3, 4 and 5 | Form 3 is initial ownership. Form 4 generally reports transactions within two business days; codes and footnotes matter. “4G” in the source is not a separate insider form. | SEC insider-transaction bulletin |
Corrections to carry through every lesson
| Source shortcut | More precise treatment |
|---|---|
| Market value or allocation change interpreted as a trade | Compare adjusted share counts and matched claims. Net additions valued at quarter end are a ranking proxy, not observed execution expenditure or motive. |
| Quarterly price range treated as manager cost basis | It is a market range, not a verified execution range. New information can reduce value after the manager bought. |
| Unexplained returns treated as proof of inefficiency | A model can omit relevant risks, use poor proxies or fit noisy data. Failure of one model does not by itself establish exploitable skill. |
| Delta neutrality or dollar neutrality treated as total risk neutrality | Zero delta is local first-order sensitivity. Dollar balance does not ensure beta balance; nonlinear, basis and liquidity risks remain. |
| CAGR over five annual observations | Five year-end observations contain four year-to-year intervals. Count actual elapsed time before choosing the exponent. |
| “IRR” as EPS divided by price | Call this earnings yield. It is not the internal rate of return on a dated sequence of cash flows. |
| Dividends discounted using WACC | Discount equity cash flows using a compatible cost of equity; WACC matches cash flows to the firm. See the Chapter 20 labs and NYU reference. |
| Terminal value discounted one extra period | If terminal value is located at the end of year N and uses year N+1 dividends, discount it N periods, not N+1. |
| Average trade returns annualized using average duration | Trades can overlap, have different capital allocations and idle cash. Construct a dated portfolio wealth path or appropriate money-weighted calculation instead. |
| Value averaging’s lower average cost treated as guaranteed superiority | Contributions and cash reserves differ. Compare terminal wealth, total resources, risk and cash-flow timing; do not compare cost basis alone. |
Seven Flexible Study Blocks
These are work blocks, not a seven-day promise. Read, calculate and revise before moving on.
| Block | Reading | Deliverable |
|---|---|---|
| 1 — Observation | Preface and Chapters 1–3 | A dated filing record, adjusted-share example and common-risk map. |
| 2 — Claims and control | Chapters 4–7 | A security-rights comparison and catalyst failure tree. |
| 3 — Visibility | Chapters 8–13 | A coverage memo comparing long-only, macro, distressed and endowment mandates. |
| 4 — Mechanical rules | Chapters 14–17 | A written ranking, weighting, boundary and rebalancing specification. |
| 5 — Allocation and disclosure | Chapters 18–19 | A no-eligible-asset policy, disclosure timeline and price-versus-value memo. |
| 6 — Business analysis | Chapter 20 and its additional labs | A ratio audit, valuation sensitivity table and timestamped thesis. |
| 7 — Portfolio judgment | Chapters 21–22; worksheets and capstone | A cash stress test, manual indicator observations and final research memo. |
Complete Chapter Map
| Chapter | Printed pages | What you will practice |
|---|---|---|
| 1. Introduction | 3–16 | Separate a disclosed holdings snapshot from a trade record and a complete portfolio. |
| 2. Bill Ackman | 17–24 | Distinguish an activist thesis from access to an activist’s influence and transaction terms. |
| 3. Bruce Berkowitz | 25–32 | Recognize common economic exposure inside a concentrated value portfolio. |
| 4. Warren Buffett | 33–40 | Separate listed stock selection from the economics of an operating holding company. |
| 5. Ian Cumming and Joseph S. Steinberg | 41–46 | Identify which parts of a turnaround investment are available to a public follower. |
| 6. David Einhorn | 47–54 | Explain why independently researched long and short positions cannot be recovered from long holdings alone. |
| 7. Carl Icahn | 55–62 | Evaluate activism as a contingent process rather than a guaranteed catalyst. |
| 8. Seth Klarman | 63–68 | Use a valuation range and recognize how little of a diversified mandate may be visible. |
| 9. John Paulson | 69–76 | Recognize timing, instruments and changing style as barriers to replication. |
| 10. Wilbur Ross | 77–84 | Distinguish a cheap distressed claim from an apparently cheap listed share. |
| 11. George Soros | 85–92 | Translate reflexivity into an observable hypothesis without pretending to observe a full macro portfolio. |
| 12. David Swensen | 93–100 | Separate endowment asset allocation from a tiny disclosed stock sleeve. |
| 13. Prem Watsa | 101–108 | Distinguish an equity thesis from the hedged insurance portfolio containing it. |
| 14. Introduction: mechanical approaches | 111–134 | Specify manager selection, security selection and allocation as separate rules. |
| 15. Equal Allocation Models | 135–148 | Implement equal allocation while defining the rebalance trigger precisely. |
| 16. Weighted Allocation Models | 149–162 | Normalize score weights and measure the concentration they create. |
| 17. Ten-Five-Two (10–5–2) Allocation Models | 163–176 | Check that tier counts and allocations produce a fully specified portfolio. |
| 18. Alternate Models | 177–192 | Separate security-cloning models from allocation and trend-filter experiments. |
| 19. Introduction: learning from the masters | 195–208 | Build a dated holdings-change worksheet without treating it as proof of intent or cost basis. |
| 20. Fundamental Analysis | 209–242 | Use compatible cash flows, discount rates and periods to test an investment thesis. |
| 21. Types of Positions and Sizing | 243–264 | Connect position sizing, diversification and contribution policy to feasible cash requirements. |
| 22. Conclusion | 265–272 | Choose a research process and define conditions that would disconfirm it. |
Part I: Tracking the Managers
Chapter 1: Introduction
Read: printed pp. 3–16. Learning objective: Separate a disclosed holdings snapshot from a trade record and a complete portfolio.
Work through the idea
A filing is an observation with a date, scope and reporting convention. A position can grow in market value because its price rises even when the share count falls. Start with security identity, share class, shares, put/call status, reporting period and public availability. Then reconcile corporate actions before interpreting changes. The book introduces several styles; their suitability for following depends on how much of the economic position you can actually observe, and how long the idea remains relevant.
Original worked example
A fictional manager reports 1,000 shares at $20 in one quarter and 900 at $25 in the next. Value rises from $20,000 to $22,500, or 12.5%, while the reported share count falls 10%. The increase in market value is not evidence of a purchase.
Even a clean share comparison gives a net change between snapshots. It cannot reveal all intervening purchases and sales. A zero closing balance does not tell you when or at what price the manager exited.
Workbook exercise 1
An unchanged 800-share position rises from $30 to $36. Calculate both values, the value change and the net share change. List three missing facts needed to copy the manager’s economic exposure.
Show the explained answer — Chapter 1
Values are $24,000 and $28,800; value rises 20%, while net shares are unchanged. Missing facts include offsetting shorts, cash and other assets, trade dates, financing and the reason for holding. An unchanged closing count also permits intervening round trips.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 2: Bill Ackman
Read: printed pp. 17–24. Learning objective: Distinguish an activist thesis from access to an activist’s influence and transaction terms.
Work through the idea
The historical Ackman profile pairs concentrated activism with examples that worked and examples that did not. A public follower can own shares without possessing board influence, negotiated rights or the capacity to finance a restructuring. Evaluate the proposed catalyst, the route by which it changes cash flows, and the cost of waiting. A position that later dominates a portfolio may have started small and appreciated; its final weight is not its original allocation.
Original worked example
Allocate 2% of a $100,000 fictional portfolio to a company. If that holding becomes worth ten times its purchase value and everything else stays flat, total wealth becomes $118,000. Its new weight is $20,000/$118,000 = 16.95%, despite no additional purchase.
A famous manager’s public involvement can itself change the available price. Buying the same issuer after disclosure does not establish the same expected payoff.
Workbook exercise 2
An activist opportunity has assumed outcomes of +30% with probability 60% and −40% with probability 40%. Find its expected simple return and the contribution from a 5% allocation if failure occurs.
Show the explained answer — Chapter 2
Expected return is 0.60×30% − 0.40×40% = 2%, before costs and time value. Failure contributes −2 percentage points to the whole portfolio if the other 95% is flat. Those subjective probabilities need evidence; a positive arithmetic expectation is not a recommendation.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 3: Bruce Berkowitz
Read: printed pp. 25–32. Learning objective: Recognize common economic exposure inside a concentrated value portfolio.
Work through the idea
This profile makes patience, liquidity and security valuation inseparable. A concentrated financial-sector portfolio may contain different companies that all depend on the same credit cycle. Investor withdrawals can force sales even when an analyst’s estimated value has not changed. Treat a falling price as a prompt to recheck asset quality, capital needs and the thesis rather than as automatic evidence of a larger bargain.
Original worked example
Three fictional financial holdings each carry a 20% portfolio weight. In a shared credit shock each loses 35%. Together they subtract 3×20%×35% = 21 percentage points from portfolio value, assuming other holdings stay flat.
Weights in the visible filing sleeve are not necessarily weights in the manager’s total assets. Multiple funds can also be aggregated under one reporting manager.
Workbook exercise 3
A $10 million fund has $1 million cash and must meet $2.5 million of redemptions. Ignoring costs, how much must it raise by selling holdings? Why might that matter to a patient investor?
Show the explained answer — Chapter 3
It must raise $1.5 million beyond its existing cash. Sales can crystallize losses or incur market impact. A long time horizon does not remove near-term liquidity obligations. A follower with different cash needs has a different feasible holding period.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 4: Warren Buffett
Read: printed pp. 33–40. Learning objective: Separate listed stock selection from the economics of an operating holding company.
Work through the idea
The Buffett chapter contrasts business ownership with purchases of marketable securities. A holding company’s insurance operations, subsidiaries, financing and negotiated preferred securities are not reproduced by following its reported common-stock positions. Read the profile for business quality and valuation discipline, then ask which claim an outside buyer can actually acquire. Book-value growth, a share-price return and an index total return measure different objects.
Original worked example
A fictional company receives a preferred investment paying an 8% contractual dividend plus warrants. An outside buyer purchases common shares yielding 2%. Even if both investors have exposure to the same business, their income, downside priority and upside rights differ.
Insurance float comes with liabilities and underwriting risk. It is not a free, permanent pool of personal investing capital. A durable business can still be bought at an excessive price.
Workbook exercise 4
A business is estimated to be worth $80 per share, but the bear-case estimate is $50. At a $60 price, calculate the discount to the base estimate and the loss to the bear-case value.
Show the explained answer — Chapter 4
The base-estimate discount is (80−60)/80 = 25%. The decline from purchase price to the bear value is (50−60)/60 = −16.67%. The first number is a valuation margin; it is not a guaranteed loss limit. Estimates can be wrong.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 5: Ian Cumming and Joseph S. Steinberg
Read: printed pp. 41–46. Learning objective: Identify which parts of a turnaround investment are available to a public follower.
Work through the idea
The Leucadia cases emphasize troubled assets, patient restructuring and transaction design. Buying a debt claim that converts into control is economically different from buying ordinary shares after the restructuring succeeds. A reporting sleeve concentrated in two companies can sit inside a much broader operating group. Assess whether the apparent concentration comes from a complete investment book or merely the reporting perimeter.
Original worked example
A fictional restructuring allocates 70% of the new equity to senior creditors, 25% to new capital and 5% to former common shareholders. Recovery in the operating business does not restore the former shareholders to their old percentage ownership.
Historical organizational names and transactions in this chapter describe the book’s period. They are not a map of the companies’ current structures.
Workbook exercise 5
A holding company has $500 million in listed securities and $2 billion of other net assets. One listed holding is $300 million. Compute its weight in the listed sleeve and in the combined $2.5 billion assumed asset base.
Show the explained answer — Chapter 5
It is 60% of the listed sleeve but 12% of the combined assumed base. These are different denominators. The calculation only works if the stated bases are comparable and measured consistently; real holding-company valuation also requires debt, liabilities and valuation adjustments.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 6: David Einhorn
Read: printed pp. 47–54. Learning objective: Explain why independently researched long and short positions cannot be recovered from long holdings alone.
Work through the idea
The Einhorn profile stresses finding the cause of mispricing and checking financial statements. A short position can express a separate thesis rather than simply hedge a named long. A long-only follower loses that part of the portfolio design. Differences in interim trimming, public disclosure and execution can also make an apparently successful manager position an unsuccessful copied position.
Original worked example
On starting capital of $100,000, a fictional portfolio has $70,000 long exposure and $30,000 short exposure. If longs fall 15% and the shorted securities fall 20%, the simple contributions are −$10,500 and +$6,000: −4.5% combined before financing and other costs.
The same long basket on its own loses 15% of its invested amount. A disclosed long allocation does not identify the manager’s actual net or gross risk.
Workbook exercise 6
An investor shorts $10,000 of stock. It rises 60%. Calculate the mark-to-market loss and explain why a stop instruction is not a guaranteed cap.
Show the explained answer — Chapter 6
The loss is $6,000 before borrow charges and other costs. Short losses can grow beyond the initial proceeds. Gaps, liquidity problems, recalls and execution slippage can prevent an exit at an intended level. A long-only clone cannot recreate that exposure by merely excluding a stock.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 7: Carl Icahn
Read: printed pp. 55–62. Learning objective: Evaluate activism as a contingent process rather than a guaranteed catalyst.
Work through the idea
The Icahn chapter supplies a useful contrast between pressure for change and the ability to obtain it. An activist proposal may meet resistance, fail commercially or arrive too late to repair a deteriorating business. Build a catalyst tree: the action sought, approvals needed, expected timing, costs and a failure case. Keep the investment thesis separate from admiration for a negotiator’s reputation.
Original worked example
A fictional target trades at $40. A successful restructuring is assumed to produce $55 of value; failure leaves $25. With an assumed 50/50 split, expected value is $40 before time and expenses. A potentially large upside is not enough by itself.
The source’s historical discussion of takeover defenses and tax rules is context, not a current legal checklist. Review actual current filings and terms for a live company.
Workbook exercise 7
Using the example, what success probability produces a $43 expected value? What additional requirement is needed before comparing that value with today’s $40?
Show the explained answer — Chapter 7
Solve 55p + 25(1−p) = 43: p = 60%. You also need the timing and an appropriate risk-adjusted valuation framework, plus transaction costs and probability justification. A future $43 and a present $40 are not directly equivalent.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 8: Seth Klarman
Read: printed pp. 63–68. Learning objective: Use a valuation range and recognize how little of a diversified mandate may be visible.
Work through the idea
The Klarman discussion emphasizes a margin of safety, holding cash when opportunities are unattractive and considering adverse scenarios. The reported common-stock sleeve can be a small part of a mandate involving distressed debt and hedges. A copied basket therefore need not reproduce either the manager’s return or protection. A margin of safety should be attached to a defensible value estimate, not inferred solely from the manager’s presence.
Original worked example
A fictional company has estimated per-share values of $35, $50 and $65 in bear, base and bull cases. At $40, the base-case discount is 20%, but the bear case still implies a 12.5% price decline.
A hedge can cost money repeatedly before a crisis occurs, and it may fail to offset the loss actually suffered. Copying selected longs omits that cost and possible payoff.
Workbook exercise 8
The public sleeve is $300 million inside $3 billion total assets; one holding is 20% of the sleeve. Calculate its assumed whole-asset weight. What would invalidate the denominator?
Show the explained answer — Chapter 8
The holding is $60 million, or 2% of the stated $3 billion base. Different measurement dates, leverage, incompatible gross/net definitions and omitted liabilities can invalidate that interpretation. Always reconcile the numerator and denominator before describing conviction.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 9: John Paulson
Read: printed pp. 69–76. Learning objective: Recognize timing, instruments and changing style as barriers to replication.
Work through the idea
The Paulson chapter shows why a prominent success in credit cannot be inferred from an equity filing. Merger spreads, credit protection, gold and mining shares have different payoff drivers. Even a disclosed position can be poor to follow if much of its prospective return has already occurred. The profile also makes style drift a research question: a manager’s advantage in one domain need not transfer to another.
Original worked example
A fictional cash acquisition offers $50 for shares trading at $48. Gross upside is $2, or 4.17%. If a failed deal leaves the stock at $34, downside is $14, or 29.17%. At an assumed 90% completion probability, the expected dollar change is 0.9×2 − 0.1×14 = $0.40 before time and costs.
Buying credit protection has premium, counterparty and contract risks. The source’s compressed “maximum loss” language is not a general rule for shorting bonds or using leverage.
Workbook exercise 9
For the example, find the completion probability that makes the expected dollar change zero. Explain why a quoted 4.17% spread is not an annualized guaranteed yield.
Show the explained answer — Chapter 9
2p − 14(1−p) = 0 gives p = 87.5%. The spread depends on completion, timing, financing and realized exit value. Annualization requires a defined holding period and does not eliminate deal-break risk. A probability just above break-even still leaves substantial uncertainty.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 10: Wilbur Ross
Read: printed pp. 77–84. Learning objective: Distinguish a cheap distressed claim from an apparently cheap listed share.
Work through the idea
The Ross profile concerns restructuring businesses and their liabilities. Where an investor sits in the capital structure determines what recovery means. A pre-IPO investment, a negotiated asset purchase and an exchange-traded share bought later can all expose investors to the same business at very different costs. A follower should trace the actual security and ownership changes rather than assume the original deal is still available.
Original worked example
A fictional business has $90 million of realizable assets and $100 million of senior claims, before other costs. Ordinary shares can be worthless even though the assets still operate and have positive value. “The business survives” is not an equity valuation.
A low quoted price compared with a historical high tells you little about the residual claim after debt, dilution and restructuring terms.
Workbook exercise 10
An early investor buys 1 million shares for $5 each. After an IPO, a follower buys 1 million at $20. Both later hold at $15. Calculate each simple return, ignoring all other terms.
Show the explained answer — Chapter 10
The early investor has a 200% gain: (15−5)/5. The follower has a 25% loss: (15−20)/20. Same company and terminal price do not mean same investment. Lockups, preferred rights and dilution can make real comparisons more complicated.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 11: George Soros
Read: printed pp. 85–92. Learning objective: Translate reflexivity into an observable hypothesis without pretending to observe a full macro portfolio.
Work through the idea
The Soros profile describes a feedback loop: prices can affect financing and behavior, which can then affect fundamentals and prices again. Turn that story into testable links rather than an all-purpose explanation after the fact. A macro portfolio may involve currencies, derivatives and fast position changes. Its reported equity and option rows are incomplete evidence of its economic exposure.
Original worked example
A fictional borrower can raise $60 against an asset valued at $100 using a 60% lending limit. If its valuation rises to $130, the same percentage permits $78 of borrowing. More financing may reinforce demand; a reversal can run the process in the other direction.
A 13F put row does not disclose enough to reconstruct the option payoff. Underlying share value is not the option premium, cash at risk or delta-adjusted bearish exposure.
Workbook exercise 11
In the example, asset value falls from $130 to $90 while debt stays $78. Calculate equity before and after. Name one missing fact before forecasting forced liquidation.
Show the explained answer — Chapter 11
Equity falls from $52 to $12, a 76.92% decline, while assets fall 30.77%. You need the loan terms, collateral tests, available liquidity and cure rights to infer forced sales. Feedback is plausible, but the price change alone does not prove its cause.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 12: David Swensen
Read: printed pp. 93–100. Learning objective: Separate endowment asset allocation from a tiny disclosed stock sleeve.
Work through the idea
The Swensen profile emphasizes horizon, equity exposure, diversification and specialist manager selection. A university’s perpetual mandate still has annual spending and cash commitments. Public holdings that represent only a small fraction of the endowment cannot describe its overall investment design. Study the allocation principles and constraints rather than interpret a large percentage within a small filing as a large percentage of the endowment.
Original worked example
An endowment of $10 billion reports a $50 million sleeve. An ETF is 80% of that sleeve: $40 million, or only 0.4% of the endowment. Copying its 80% sleeve weight would greatly change the economic role.
The source criticizes some passive allocations based on a severe bear period. A drawdown alone does not establish that a diversified long-horizon allocation is unsuitable; the mandate, spending and full return record matter.
Workbook exercise 12
An institution must pay $120 million in spending and $80 million in capital calls, with $150 million liquid cash. Calculate its funding gap. Why does a perpetual horizon not solve it?
Show the explained answer — Chapter 12
The immediate gap is $50 million. Obligations arrive on dates regardless of the institution’s long life. Asset sales, other inflows or arranged financing may be needed. A permanent institution can encounter a short-term liquidity problem.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 13: Prem Watsa
Read: printed pp. 101–108. Learning objective: Distinguish an equity thesis from the hedged insurance portfolio containing it.
Work through the idea
The Watsa profile combines underwriting discipline, long-horizon value investing and protection against adverse macro outcomes. A stock-only follower inherits neither the insurance operation nor all of the hedges. When a manager reports a large losing equity holding, ask how the full portfolio behaved and why the original thesis persists. Neither patience nor founder management proves that a declining business will recover.
Original worked example
A fictional $1 million portfolio has a $200,000 equity basket that loses 40%, while a separate hedge gains $50,000. Together those positions lose $30,000, or 3% of the whole portfolio, assuming the rest is flat. A $200,000 unhedged copy loses $80,000.
Hedge behavior, basis, cost, expiration and counterparty terms matter. A statement that a portfolio is “fully hedged” does not establish protection against every kind of loss.
Workbook exercise 13
Protection costs 2% of portfolio value each year for three years, with no payout. On a fixed $100,000 notional, what is the simple cumulative cost? What makes that different from proving it was a mistake?
Show the explained answer — Chapter 13
The fixed-notional cost is $6,000, excluding compounding and financing. Insurance can expire unused while still serving an ex-ante risk purpose. Assess the probability, affordability and relevance of the covered event, rather than judging only the realized outcome.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Part II: Mechanical Approaches
Chapter 14: Introduction: mechanical approaches
Read: printed pp. 111–134. Learning objective: Specify manager selection, security selection and allocation as separate rules.
Work through the idea
The book’s second part moves from anecdotes to reproducible portfolio instructions. It selects managers, ranks existing holdings or net additions, and allocates capital. These are separate choices with separate failure modes. Large weights are an observable ranking input, not a direct measurement of current conviction. An apparent new position requires identity and corporate-action checks; an increase in shares requires a consistently valued ranking measure.
Original worked example
In a fictional $1 million eligible sleeve, a stock rises from 10,000 to 12,000 shares after adjustment. At the quarter-end price of $30, the net-addition proxy is 2,000×30 = $60,000, or 6% of the sleeve. It is not the manager’s $60,000 expenditure: execution prices are unknown.
Historical studies and the author’s chosen successful managers motivate research. They do not remove hindsight selection, survivorship, parameter search or implementation costs.
Workbook exercise 14
Design a five-line research rule: eligible managers, eligible securities, ranking input, portfolio weighting and first executable decision time. Include how missing filings are handled.
Show the explained answer — Chapter 14
One acceptable design freezes managers before evaluation; permits identified long common shares only; ranks adjusted share additions valued at period end; equal-weights five names with a preset cap; and waits until the actual filing is public before the next feasible session. A missing filing creates no new order until the preset policy is satisfied. This is an original specification, not a validated model.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 15: Equal Allocation Models
Read: printed pp. 135–148. Learning objective: Implement equal allocation while defining the rebalance trigger precisely.
Work through the idea
Equal weighting makes allocation transparent, but it can concentrate a small model far more than the source book. The source’s three-stock demonstrations should be read as teaching examples. Separate replacement trades from rebalancing trades. Also distinguish the book’s annually refreshed new-addition example from its quarterly holdings updates. A return measured over a short favorable window does not validate all those choices.
Original worked example
A $60,000 fictional three-stock model starts at $20,000 per holding. Before review the values are $30,000, $20,000 and $18,000. Largest/smallest = 1.667, above a 1.5 trigger. An equal rebalance targets $68,000/3 = $22,666.67 each, before costs.
“50% greater than the smallest” means largest > 1.5×smallest, not largest > 0.5×smallest. The source wording varies; choose and record one strict or inclusive boundary in advance.
Workbook exercise 15
Values are $27,000, $24,000 and $18,000. Under a strict greater-than-1.5 rule, is rebalancing triggered? Under an at-least-1.5 rule? Explain what remains to be specified.
Show the explained answer — Chapter 15
The ratio is exactly 1.5: strict greater-than does not trigger; at least does. Specify the observation schedule, execution prices, costs, rounding and replacement policy too. A precise boundary prevents changing the rule after viewing its return.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 16: Weighted Allocation Models
Read: printed pp. 149–162. Learning objective: Normalize score weights and measure the concentration they create.
Work through the idea
The weighted examples map a disclosed allocation into a score, then divide each score by the sum. This is a coarse approximation, not exact replication. A small movement across a threshold can cause a large target-weight change. When combining managers, note that their reporting sleeves may differ in coverage: the same score does not guarantee the same economic conviction. Rebalance frequency and turnover become part of the model.
Original worked example
For a fictional $72,000 basket, scores 3, 2 and 1 total 6. Targets are 50%, 33.33% and 16.67%, corresponding to $36,000, $24,000 and $12,000. A three-point score does not mean a 3% position.
The book’s intervals leave some boundary wording open. For our practice only: less than 10% gets 1; 10% through 20% inclusive gets 2; above 20% gets 3. This convention resolves equality, without claiming the historical tests used it.
Workbook exercise 16
Three scores change from 2,2,1 to 3,2,1. Find the first holding’s old and new target weights. If its source weight merely crossed 20%, what risk does this illustrate?
Show the explained answer — Chapter 16
The old target is 2/5 = 40%; the new target is 3/6 = 50%. A slight source change can create a 10-percentage-point target jump. Such threshold effects can increase trading, concentration and sensitivity to data errors. A different scoring rule must be evaluated separately.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 17: Ten-Five-Two (10–5–2) Allocation Models
Read: printed pp. 163–176. Learning objective: Check that tier counts and allocations produce a fully specified portfolio.
Work through the idea
The chapter’s tiered model tries to retain important positions while simplifying a larger portfolio. The label can mean literal percentages in a compatible tier design or proportional scores in an adapted small model. Those interpretations are different. Every implementation must define the number of names per tier, ranking, treatment of ties, missing names and rebalancing. Simplifying many positions also changes the coverage of the source sleeve.
Original worked example
Three positions at 10%, twelve at 5% and five at 2% total 30%+60%+10%=100% across twenty names. By contrast, applying scores 10,5,5,2,2 to five names requires division by 24: the largest becomes 41.67%.
Tier shifts can require disproportionate trading when two nearly equal source holdings exchange ranks. More names may reduce single-name concentration while leaving a shared sector exposure intact.
Workbook exercise 17
A model uses two names at 10%, eight at 5% and fifteen at 2%. Calculate the total allocation. Provide two transparent ways to handle the remainder before any research is run.
Show the explained answer — Chapter 17
Total allocation is 20%+40%+30%=90%. One policy retains 10% cash; another renormalizes each specified weight by dividing by 0.90. The latter makes a nominal 10% position 11.11%. Neither policy can be silently treated as the other.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 18: Alternate Models
Read: printed pp. 177–192. Learning objective: Separate security-cloning models from allocation and trend-filter experiments.
Work through the idea
The alternatives include normalizing retained holdings, static multi-asset allocation, relative-strength rotation and an SMA-based filter. Each asks a different question. Removing tiny positions and rescaling the rest approximates the visible sleeve; it does not recover the full fund. A trend filter can change concentration and spend long periods lagging a rebound. The source’s short historical comparison is evidence about that sample, not a promise of positive returns in all markets.
Original worked example
Retain fictional source weights of 40%, 30% and 20%, dropping a 10% remainder. Normalize by 90% to obtain 44.44%, 33.33% and 22.22%. In a separate six-asset trend experiment with only two eligible assets, equal distribution means 50% in each, not one-sixth in each.
The source does not fully resolve what to do if no asset passes its SMA rule. Predetermine a cash policy rather than inventing an eligible asset. Also check fund mandates: an all-world ex-US fund may already contain emerging-market exposure.
Workbook exercise 18
At review, three assets close at 110, 98 and 103 with their respective 200-day SMAs at 100, 100 and 105. With a strict above-SMA rule, identify the eligible set. What concentration results if all capital is distributed among it?
Show the explained answer — Chapter 18
Only the first asset qualifies, so an unrestricted equal-among-eligible model assigns it 100%. That is a new concentration risk. A cap with cash remainder would be a different, explicitly stated rule. The chart indicator below observes one asset; it does not allocate among a universe.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Part III: Learning to Assess Independently
Chapter 19: Introduction: learning from the masters
Read: printed pp. 195–208. Learning objective: Build a dated holdings-change worksheet without treating it as proof of intent or cost basis.
Work through the idea
The third part adds price comparisons and other disclosure forms. A quarterly high-low range describes available market prices during that period, not verified manager executions. A price below that range can reflect new bad information. Use net share changes and matched security identity instead of changes in portfolio percentages alone. Distinguish ownership of an issuer from weight in a manager’s assets; those are separate ratios.
Original worked example
A fictional manager increases adjusted shares from 100,000 to 120,000. At a $25 period-end price, the net-addition proxy is $500,000. If the eligible sleeve is $20 million, that proxy is 2.5% of the sleeve. A later $22 quote is lower, but gives no proof that business value is unchanged.
Use the current SEC timing note above. Schedule 13G has filer-dependent deadlines; it is not universally a fast version of 13D. Form 3 describes initial ownership, not necessarily a fresh purchase.
Workbook exercise 19
A filing shows Form 4 transaction code A, another shows P, and a third shows F. Should they all be ranked as open-market insider purchases? What records should accompany a proposed signal?
Show the explained answer — Chapter 19
No. A denotes an award or other acquisition from the company; P covers a purchase on an exchange or from another person; F covers shares used for exercise price or tax liability. Inspect footnotes, prices, direct/indirect ownership, transaction and public timestamps, and related transactions. Even P is not a complete statement of investment motive.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 20: Fundamental Analysis
Read: printed pp. 209–242. Learning objective: Use compatible cash flows, discount rates and periods to test an investment thesis.
Work through the idea
This long chapter moves from a qualitative checklist through profitability, liquidity, financial-strength scores and several valuation models. Treat each metric as a question rather than a verdict. High ROE can reflect a small equity denominator or leverage; low P/B can reflect impaired assets. Screens such as F-score and Z-score have model and sector limits. A valuation range depends on reinvestment, growth, risk and capital structure. An imperfect factor model does not logically prove an exploitable inefficiency.
Original worked example
For a fictional stable dividend payer, next year’s dividend is $3, the required equity return is 10% and perpetual growth is 4%. Gordon value is 3/(0.10−0.04) = $50. At a $40 price, the discount is 20%. Increasing the required return to 12% reduces value to $37.50, turning the apparent bargain into a price above this estimate.
The additional valuation labs below are part of this chapter’s work. They correct period counting, discount-rate matching and terminal-value timing rather than reproducing the source’s IBM tables.
Workbook exercise 20
Value the same $3 dividend at required returns of 9%, 10% and 11%, each with 3% perpetual growth. Find the valuation range and explain why EPS divided by a Treasury yield cannot establish a stock is safe.
Show the explained answer — Chapter 20
Values are $50.00, $42.86 and $37.50, a $37.50–$50 range under those assumptions. Corporate earnings are risky, are not necessarily distributable and may shrink. A Treasury yield alone omits equity risk, reinvestment and capital structure. “Cheap relative to a risk-free capitalization” is not a safety conclusion.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Continue with the additional Chapter 20 financial-analysis labs before completing this chapter.
Chapter 21: Types of Positions and Sizing
Read: printed pp. 243–264. Learning objective: Connect position sizing, diversification and contribution policy to feasible cash requirements.
Work through the idea
The chapter joins covariance, cash reserves, staged purchases and small asymmetric bets. Correlations estimated from past returns can change when liquidity disappears. Zero correlation does not mean independence, and a normal-distribution illustration is not a reliable map of tail losses. Dollar-cost averaging fixes contributions; share averaging fixes shares; value averaging fixes a target path and can demand large cash additions or sales. Compare wealth and external cash flows, not cost basis alone.
Original worked example
With fictional prices of $20 then $10, investing $1,000 each period buys 50+100 = 150 shares for $2,000: average cost $13.33. Buying 50 shares each period costs $1,500 for 100 shares: average $15. These methods deploy different capital, so the lower cost basis alone does not establish a superior investment result.
Cash has opportunity and inflation costs, but interest-bearing cash instruments can earn income. The “cash as an option” description is a metaphor, not a way to price cash from a call’s time value. A small position is only loss-bounded if its legal payoff and financing actually bound losses.
Workbook exercise 21
A value-averaging plan targets $5,000 today. The existing holding is worth $3,200. Calculate the required contribution. Repeat if it is worth $5,700. Explain how a $1,000 contribution cap changes the plan.
Show the explained answer — Chapter 21
The first case requires a $1,800 purchase; the second a $700 sale. A $1,000 cap leaves the first case $800 short of target, so the investor is following a modified rule. The policy needs a cash reserve and a response to funding shortfalls; lower prices do not create unlimited available capital.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 22: Conclusion
Read: printed pp. 265–272. Learning objective: Choose a research process and define conditions that would disconfirm it.
Work through the idea
The conclusion distinguishes selecting individual disclosed positions, constructing mechanical baskets and learning to assess businesses independently. The strongest practical outcome is a process you can explain, date and audit. Favoring an observable slow-moving mandate can reduce replication problems but does not guarantee returns or remove the need for portfolio limits. A decision to keep a diversified benchmark allocation while researching a smaller sleeve is also a valid outcome.
Original worked example
A fictional research sleeve is capped at 15% of a $200,000 portfolio: $30,000. With a 10% single-name cap within that sleeve, maximum initial capital per name is $3,000, or 1.5% of the whole portfolio. State clearly which denominator a limit uses.
A short backtest, a selected winner or a high estimated value is insufficient grounds to scale. Keep the manager universe and rules frozen before examining a genuinely later evaluation period.
Workbook exercise 22
Write three reasons to pause a cloning experiment and distinguish a data failure from an investment loss. Set a review interval and a benchmark before opening the answer.
Show the explained answer — Chapter 22
Examples: pause for unreconciled corporate actions or timestamps; pause when concentration or cash limits are breached; review when costs or actual coverage materially differ from assumptions. A valid investment can lose money; corrupt data cannot support its test. Use a preset review date and an investable benchmark with comparable risk and return accounting. The capstone asks you to document these decisions.
Return to the reading: record the relevant printed page, identify one assumption or limitation, and explain how it affects your answer. Use your own words rather than copying the source.
Chapter 20: Additional Financial-Analysis Labs
The chapter spans many methods. Use these original drills to check the mechanics behind them. These are simplified examples, not screens for buying a stock. Define each ratio before comparing data from different providers.
Lab A: Build a coherent ratio sheet
Fictional annual net income is $12 million, EBIT $18 million, average equity $60 million, average assets $150 million, current assets $45 million and current liabilities $30 million. Cash is $8 million, short-term investments $2 million, receivables $14 million and capital employed $100 million. Assume all figures cover compatible periods and definitions.
| Metric | Calculation | Interpretation limit |
|---|---|---|
| ROE | 12/60 = 20% | High leverage or buybacks can raise it without improving the underlying business. |
| ROA, net-income convention | 12/150 = 8% | State whether assets are averaged. Compare compatible industries and periods. |
| ROCE, our EBIT convention | 18/100 = 18% | Definitions differ; do not silently equate this with the book’s net-income-based expression. |
| Current ratio | 45/30 = 1.50 | Asset quality and liability timing matter; a threshold is not a solvency guarantee. |
| Quick ratio, stated liquid-asset convention | (8+2+14)/30 = 0.80 | Do not count uncollectible receivables as cash. |
| Cash conversion cycle | DIO + DSO − DPO | Use compatible inventory, receivable, payable and flow definitions. |
Exercise A: DIO is 50 days, DSO 35 and DPO 40. Find the cash conversion cycle. If receivables rise while reported sales remain flat, what should you investigate?
Show explained answer — Lab A
CCC = 50+35−40 = 45 days. Investigate collections, sales terms, revenue recognition, customer quality and seasonality. An increase is a prompt for analysis, not proof of fraud. The book’s simple threshold rules should be interpreted in a sector and business-cycle context.
Lab B: Screens are filters, not valuation conclusions
| Tool | Research question | Required caution |
|---|---|---|
| PEG | Does the quoted P/E look high relative to the chosen earnings-growth estimate? | Growth must be defined. P/E 18 divided by growth 12 (percentage points) = 1.5; dividing by 0.12 gives a different, inappropriate scale for this convention. Negative earnings/growth can make the shortcut unhelpful. |
| Price/book | What is the market paying relative to accounting equity? | Write-downs, intangible assets, leverage and negative book equity can dominate the ratio. |
| Earnings yield | What is normalized EPS divided by price? | EPS is not necessarily available cash or a shareholder payout. It is not internal rate of return. |
| Piotroski F-score | Do the specified financial-strength signals improve? | Choose the precise implementation, fiscal years and accounting inputs; an attractive score does not establish intrinsic value. |
| Altman Z-score | Does a particular distress model flag financial vulnerability? | Choose the version for the business and accounting setting. A zone is not a guaranteed bankruptcy prediction or an all-sector safe label. |
| Graham number and net-current-asset screens | Does a stock meet a particular conservative accounting-based screen? | Check earnings and book-value applicability, realizable asset values, all senior claims, dilution and the original qualitative conditions. |
Exercise B: Price is $48, EPS is $3, expected annual earnings growth is 8%, and book value per share is $12. Calculate P/E, PEG under the stated convention, P/B and earnings yield. Does any one prove undervaluation?
Show explained answer — Lab B
P/E = 16; PEG = 16/8 = 2; P/B = 4; earnings yield = 3/48 = 6.25%. None proves fair value. Growth is uncertain, accounting earnings differ from distributions, and comparable firms may have different risks, reinvestment needs and asset quality.
Lab C: Count growth periods correctly
EPS rises from $2 at the end of Year 1 to $3 at the end of Year 5. Those are four elapsed annual intervals, so CAGR = (3/2)1/4−1 = 10.67%. Using 1/5 would give 8.45%, which describes a different time interval. The source’s 2007-to-2011 examples should be checked using actual elapsed dates, rather than the count of annual labels.
Exercise C: Dividends rise from $1.00 at the end of 2020 to $1.46 at the end of 2024. Compute the four-year compound rate. Explain why the next four years need not repeat it.
Show explained answer — Lab C
(1.46/1.00)^(1/4)−1 ≈ 9.92%. This is a historical rate. Future distributions depend on earnings, reinvestment, payout policy, capital structure and changing business conditions. Historical compounding is not a forecast.
Lab D: Value equity with a dated dividend timeline
Our fictional company pays $2 at the end of Year 1, $2.20 at the end of Year 2 and $2.40 at the end of Year 3. Thereafter dividends grow 3% annually. Required equity return is 10%. The Year 4 dividend is $2.472, so terminal value at the end of Year 3 is 2.472/(0.10−0.03) = $35.3143. Present value is 2/1.10 + 2.20/1.10² + (2.40+35.3143)/1.10³ = $31.97.
Use the cost of equity for dividends, not WACC. Terminal value calculated with the next year’s dividend is located one period before that dividend: discount it from Year 3 here, not Year 4. The source’s two-step example uses a terminal-value discount exponent that should be checked against its stated timeline. NYU valuation notes explain cash-flow/rate matching; NYU terminal-value notes define the terminal claim.
Exercise D: Using these same dividends, recalculate with 11% required return and 3% perpetual growth. Why is terminal value a major source of uncertainty?
Show explained answer — Lab D
Terminal value at Year 3 = 2.472/(0.11−0.03) = $30.90. Present value = 2/1.11 + 2.20/1.11² + (2.40+30.90)/1.11³ ≈ $27.94. Most value comes from cash flows after the explicit forecast, so a modest discount-rate or growth assumption change can shift the estimate materially. Growth must remain below the required return in this stable-growth model.
Lab E: Reconcile a non-dividend and liquidation case
For a non-dividend payer, specify the first actual dividend date. If the first dividend is in Year 6, the Gordon value using that dividend lies at the end of Year 5; discount that claim five periods, while accounting for any earlier payouts separately. Alternatively model cash flow to equity with compatible assumptions. Do not add EPS as if it were an extra distribution on top of dividends.
For an original liquidation screen, assume current assets of $40 million, total liabilities of $28 million and 4 million shares, with no other senior claims. NCAV/share = ($40m−$28m)/4m = $3.00; a two-thirds purchase screen is $2.00. With cash $10m, receivables $12m and inventory $18m, the book-style NNWC haircut gives ($10m+0.75×$12m+0.5×$18m−$28m)/4m = $0.00. Different assumptions change the screen radically.
Exercise E: Why is the two-thirds NCAV threshold not the same thing as estimated liquidation value? Name three items that could reduce actual recovery further.
Show explained answer — Lab E
NCAV uses accounting current assets less total liabilities, and the two-thirds multiplier is a purchase-screen discount. Neither is a binding liquidation distribution. Collection shortfalls, obsolete inventory, winding-up expenses, contingent liabilities, taxes and additional senior claims can lower actual recovery. Some businesses burn value before liquidation occurs.
Compare the Models Before Testing
| Approach | Main input | Main research cost |
|---|---|---|
| Ad hoc disclosed position | One dated holding plus an independent thesis | Portfolio fit and exits can be arbitrary unless specified. |
| Equal-weighted top holdings | Latest eligible holdings ranked by value | A small basket amplifies single-name and sector risks. |
| Largest net additions | Two comparable snapshots and adjustment records | More noisy changes, missing history and potentially higher turnover. |
| Score-weighted basket | A defined mapping from source weights to scores | Threshold discontinuities and required rebalancing. |
| Tiered allocation | Ranks, tier counts and weights that add to 100% | Rank churn and imperfect sleeve coverage. |
| Retained-weight approximation | Eligible holdings after cutoff, normalized | It remains a clone of the visible reporting slice. |
| Static multi-asset allocation | Mandates and predetermined target weights | Drawdowns, changing correlations and rebalancing costs. |
| Rotation or SMA overlay | Dated prices, observation schedule and eligibility rule | Whipsaw, concentration, idle cash and execution delay. |
Minimum rules for a credible comparison
- Freeze manager membership and selection criteria before the evaluation period; retain dead, renamed and delisted candidates as appropriate.
- Store original filings, amendments, acceptance timestamps, reporting periods and schema units. Do not let later corrections leak into earlier decisions.
- Use security-level identity, share classes, option flags and corporate-action adjustments. Document affiliate aggregation and missing rows.
- Choose price versus total-return accounting consistently. Include dividends, delisting proceeds, splits, cash yield, spreads, fees and plausible impact.
- Use only information available before execution. A public after-close filing cannot authorize a fill at that same close.
- Define ranking ties, threshold boundaries, position caps, no-candidate policy, rebalance cadence and cash handling before testing.
- Report benchmark-relative and absolute outcomes, drawdown, turnover and concentration over an untouched period. Test alternative start dates and costs without selecting only favorable results.
Original cost exercise: A $50,000 model replaces 40% of its value. It sells $20,000 and buys $20,000, with a hypothetical cost of 0.25% of each traded dollar. Cost is $100, or 0.20% of portfolio value. A zero commission schedule would not eliminate spreads or impact. Those cost assumptions must be measured or justified; they are not market quotes.
Pine Indicator and Manual Observation Lab
What This Indicator Shows
Chapter 18’s 200-day SMA discussion motivates this original, deliberately narrow indicator. On a standard 1D candlestick chart, it plots a configurable SMA (default 200 daily closes), marks a confirmed transition above, below or exactly equal to it, and shades the last confirmed regime. The first valid close initializes state without a transition marker. A later transition produces one marker; remaining in that regime does not repeat it.
| Item | Definition | Practical limit |
|---|---|---|
| Input | One standard daily close series; default 200-bar SMA. | Daily chart bars are trading observations, not calendar days. Price adjustments and feeds can differ. |
| Above/below/equal | At a confirmed close, compare close with the contemporaneous SMA using strict > and <; equality is neutral. | The SMA includes that close. Exact equality may be rare. |
| State | Last confirmed regime persists until the next valid confirmed close changes it. | No expiring setup, order replacement, stop, target or broker fill is modeled. Invalid data resets initialization. |
| Live values | The SMA and distance percentage can change during an open bar. | Markers and regime state wait for confirmation; the shaded state remains the last confirmed observation. |
| Unsupported timeframe | Hide signals and display a 1D instruction outside a one-day chart. | Synthetic charts such as Heikin Ashi still require the reader to switch to standard candles. |
| Alerts | One event condition per confirmed regime transition. | Create alerts with once-per-bar-close settings. Alert delivery is not an execution guarantee. |
What This Indicator Does Not Show
It does not import EDGAR filings, identify hedge fund holdings or institutional intent, infer manager trades, create a multi-asset portfolio, or reproduce the book’s quarterly allocation test. It observes a single chart daily. It has no simulated orders and establishes no returns, drawdown, win rate, fill quality or profitability. The indicator does not label a price decline as undervaluation.
Validation status
The script received static checks for complete code, indicator-only declarations, daily gating, confirmation timing and exact text preservation in the saved draft. It has not been compiled in TradingView or tested on a chart. The copy interface has been structurally checked, not browser-tested. Confirmed markers reduce intrabar changes, but revised historical data, settings and chart adjustments can still alter history. There are no higher-timeframe requests or lookahead settings.
View Pine Script Indicator Code
//@version=6
indicator("CandleChasers Confirmed SMA Regime Lab", overlay=true)
length = input.int(200, "SMA length (daily bars)", minval=2)
showBackground = input.bool(true, "Shade confirmed regime")
isDaily = timeframe.isdaily and timeframe.multiplier == 1
sma = ta.sma(close, length)
ready = isDaily and not na(sma) and not na(close) and close > 0
// Regime changes only on confirmed chart closes. Equality is the neutral state.
var int regime = 0
var bool initialized = false
bool aboveEvent = false
bool belowEvent = false
bool equalEvent = false
if barstate.isconfirmed
if ready
int nextRegime = close > sma ? 1 : close < sma ? -1 : 0
if initialized
aboveEvent := nextRegime == 1 and regime != 1
belowEvent := nextRegime == -1 and regime != -1
equalEvent := nextRegime == 0 and regime != 0
regime := nextRegime
initialized := true
else
regime := 0
initialized := false
plot(isDaily ? sma : na, "Daily SMA", color=color.blue, linewidth=2)
plot(isDaily and initialized ? regime : na, "Confirmed regime (+1 / 0 / -1)", display=display.data_window)
plot(isDaily and ready ? 100.0 * (close / sma - 1.0) : na, "Live distance from SMA (%)", display=display.data_window)
plotshape(aboveEvent, title="Confirmed above", text="ABOVE", style=shape.triangleup, location=location.belowbar, color=color.teal, textcolor=color.white, size=size.tiny)
plotshape(belowEvent, title="Confirmed below", text="BELOW", style=shape.triangledown, location=location.abovebar, color=color.orange, textcolor=color.black, size=size.tiny)
plotshape(equalEvent, title="Confirmed equal", text="EQUAL", style=shape.circle, location=location.abovebar, color=color.gray, size=size.tiny)
color shade = not initialized or not ready ? na : regime == 1 ? color.new(color.teal, 92) : regime == -1 ? color.new(color.orange, 92) : color.new(color.gray, 92)
bgcolor(showBackground ? shade : na, title="Last confirmed regime")
alertcondition(aboveEvent, "Confirmed above SMA", "Daily close entered the above-SMA regime. Observation only.")
alertcondition(belowEvent, "Confirmed below SMA", "Daily close entered the below-SMA regime. Observation only.")
alertcondition(equalEvent, "Confirmed equal to SMA", "Daily close equaled the SMA. Observation only.")
var table status = table.new(position.top_right, 1, 1)
if barstate.islast
string label = not isDaily ? "Use a standard 1D chart" : not ready ? "Waiting for SMA warmup" : not initialized ? "Waiting for confirmed close" : regime == 1 ? "Last confirmed: ABOVE" : regime == -1 ? "Last confirmed: BELOW" : "Last confirmed: EQUAL"
table.cell(status, 0, 0, label, text_color=color.white, bgcolor=color.new(color.black, 20))
Manual replay protocol
- Choose a liquid listed instrument, standard 1D candles and a fixed observation window. Record symbol, feed, adjustment setting, SMA length and dates before looking for favorable episodes.
- Allow at least the required warmup. Independently average the first 200 eligible closes in a spreadsheet and compare with the chart’s first valid SMA.
- Step through confirmed closes. Record the prior regime, close, SMA, new regime and whether a marker should appear. A first valid observation initializes without a marker.
- Check a persistent above stretch, an above-to-below switch, a later recovery and any equality case. Observe an open daily bar to distinguish live SMA movement from confirmed state.
- Compare daily observation with a separately specified quarterly review schedule. Do not present the two as the same rule or use the day’s final close before it exists.
- Record all observations, including quick reversals. Any hypothetical trade comparison must separately specify next feasible execution, costs, cash returns and risk limits.
Original state exercise: After initialization, confirmed close-minus-SMA differences are +2, +1, −1, −3, 0 and +1. Starting from the above regime, identify the transition markers.
Show explained answer — Indicator state exercise
The first two closes remain above, so no new marker. The third produces BELOW; the fourth remains below. The fifth produces EQUAL and the sixth ABOVE. These are observations, not entry and exit fills. The first valid bar would not produce a transition marker if it had been the initialization bar.
Four Reusable Worksheets
Copy these tables into your own notes or spreadsheet. The response column tells you what to record; it does not save entries on this page. Complete the evidence and timing fields before calculating a proposed allocation.
1. Filing and observability audit
| Field | Record this |
|---|---|
| Manager identity and aggregation | Legal filer, CIK, affiliated reporters and double-count policy. |
| Reporting quarter and source accession | Quarter end, original filing link, amendment links and archive date. |
| Public timestamp and first feasible review | Acceptance time, timezone, market session and assumed execution delay. |
| Schema units and eligible security types | Dollars versus thousands, CUSIP/class, put/call flags and share/principal units. |
| Coverage and exclusions | Which sleeve is visible; what cash, shorts or other exposures remain unknown. |
| Adjusted-share reconciliation | Prior/current shares, split factors, spin-offs, identity changes and unexplained gaps. |
| Amendments and missing information | What was knowable at the decision time; what remains unresolved. |
2. Model and rebalance specification
| Field | Record this |
|---|---|
| Eligible universe | Manager membership frozen on a stated date; securities allowed and excluded. |
| Ranking and boundaries | Largest holding or adjusted net addition; score thresholds, ties and equality. |
| Weight formula and caps | Target formula, denominator, whole-portfolio/sleeve limits and sum-to-100 check. |
| Review schedule | Observation and execution dates; replacement versus drift rebalancing. |
| No candidate or missing filing | Precommitted hold/cash rule and when unresolved data blocks a new decision. |
| Orders, cash and costs | Pretrade values, exact dollars, share rounding, spreads, fees and cash yield. |
| Evaluation plan | Untouched dates, investable benchmark, total-return accounting and failure criteria. |
3. Independent business and valuation memo
| Field | Record this |
|---|---|
| Claim and business | Exact security rights, revenue mechanism and why this falls within your competence. |
| Evidence and counterevidence | Dated financial reports, original thesis and evidence against it. |
| Balance sheet and liquidity | Senior claims, dilution, cash conversion, funding dates and refinancing stress. |
| Valuation range | Bear/base/bull assumptions, compatible discount rate, periods and sensitivity. |
| Catalyst and timing | Mechanism, dependencies, time to resolution and failure path. |
| Portfolio fit | Common drivers, existing exposures, contribution under a loss and cash requirements. |
| Decision and review trigger | Proceed/research/defer/reject; maximum size and what would change your view. |
4. Replay and research journal
| Field | Record this |
|---|---|
| Precommitment | Symbol, settings, fixed dates, hypothesis and comparison rule before inspection. |
| Observation row | Date, confirmed close, SMA, old/new state, event and relevant filing availability. |
| Signal timing | What was known before the decision; after-close or later amendments excluded. |
| Execution assumption | Next feasible price, costs, cash handling and hypothetical status. |
| Adverse evidence | Whipsaws, concentration, missing records, outliers and alternative explanations. |
| Full-period results | All episodes, benchmark, drawdown path, turnover and uncertainty; no cherry-picking. |
| Next decision | Retain, revise for a new test, or abandon; preserve the previous rule and result. |
Capstone: A Disclosure-to-Decision Research Memo
Your deliverable is a two-to-four-page memo plus a calculation sheet, not a recommendation to invest real money. Use the fictional dataset below for the calculations, then use the book to defend your process. Finish all sections before opening the key.
| Item | Fictional facts |
|---|---|
| Portfolio | Total $200,000. Research sleeve maximum 15%. All other assets stay flat in the simple stress exercise. |
| Disclosure | Quarter ends June 30. Filing becomes public August 12 after the market closes. No earlier disclosure of these holdings is assumed. |
| Company A | Adjusted shares rise from 100,000 to 110,000; quarter-end price $40; eligible manager sleeve $20 million. |
| Company B | Adjusted shares fall from 100,000 to 90,000; price rises from $20 to $25. |
| Retained source holdings | A, C and D have source weights 30%, 20% and 10%, respectively. An omitted remainder is 40%. |
| Valuation of A | First annual dividend $2 next year; stable growth 3%; required equity return 9%; current price $30. |
| Risk policy | At most 10% of research sleeve in any one company; stress A with a 50% decline. |
| Cost test | For a separate $30,000 model, sell $12,000 and buy $12,000 at assumed 0.25% of dollars traded. |
- Draw the information timeline and state the earliest assumed feasible trade. Identify why June 30 prices cannot be execution prices for a disclosure-driven trade.
- Calculate A’s net-addition proxy and its fraction of the manager sleeve. Calculate both changes for B: shares and market value.
- Normalize the retained holdings. Identify how applying the result to the entire sleeve would interact with the single-name cap; state an explicit cash policy or reject the design.
- Calculate A’s stable-growth value and valuation margin. Revalue using 11% required equity return. Separate estimated value from loss protection.
- Calculate maximum capital in A, its whole-portfolio weight and the assumed portfolio contribution from a 50% fall.
- Calculate the cost test. List at least five missing inputs for a credible historical comparison.
- Use the book to compare at least three managers whose mandates differ in observability. End with a research/defer/reject decision and three conditions that would change it.
Show explained answer — Capstone
The first assumed opportunity is the next feasible session after August 12’s after-close disclosure, not the June 30 close or August 12 close. A adds 10,000 adjusted shares; 10,000×$40 = $400,000, or 2% of the $20 million sleeve. This is not actual expenditure. B’s shares fall 10%, while market value rises from $2 million to $2.25 million, or 12.5%. Retained weights total 60%; normalized targets are 50%, 33.33% and 16.67%. Applying these without limits conflicts with the stated 10% per-name sleeve cap; one original policy caps each at 10% and holds the remaining 70% in cash, without pretending that is the original model. A’s equity value is 2/(0.09−0.03) = $33.33; margin at $30 is 10%. At 11%, value is $25, below price. Maximum sleeve = $30,000; maximum A = $3,000 = 1.5% of total assets. A 50% decline subtracts $1,500, or 0.75 percentage points, under the stated flat-other-assets assumption. Costs = ($12,000+$12,000)×0.0025 = $60, or 0.20% of that $30,000 model. Missing inputs include actual timestamps, adjustments, dividends, delistings, costs, cash yield, frozen manager membership and a comparable benchmark. A long-horizon listed-stock mandate is more observable than an intraperiod macro book or a private restructuring, but no mandate is automatically profitable to follow.
| Criterion | Points | Full-credit standard |
|---|---|---|
| Source and information timeline | 20 | Correct quarter/public/execution dates and distinction between historical source and current rules. |
| Data and arithmetic | 20 | Adjusted shares, denominators, normalized weights, valuation, risk and cost calculations reconcile. |
| Model specification | 20 | Explicit caps, cash remainder, ties, timing, missing data and rebalance rules. |
| Independent thesis and risk | 20 | Counterevidence, valuation sensitivity, omitted exposures and funding constraints. |
| Research judgment | 20 | Fair benchmark, later evaluation period, honest uncertainty and actionable stop/review conditions. |
A score of 80/100 is a useful learning target, not permission to trade. Any unresolved timestamp, security-identity or unit error requires correction regardless of the score.
Technical and Current-Context References
The supplied book is the reading source. The following primary materials support the narrow technical updates; they do not validate the book’s returns or this workbook’s experimental models.
- SEC: Frequently Asked Questions About Form 13F — timing, units, missing exposure and options.
- SEC: Form 13F instructions — scope, shared reporting and information-table definitions.
- SEC: Beneficial ownership reporting interpretations — current 13D/13G context.
- SEC: Insider Transactions and Forms 3, 4 and 5 — form purposes and transaction codes; consult current form instructions for compliance.
- Aswath Damodaran, NYU: Valuation notes — equity versus firm cash flows and discount rates.
- Aswath Damodaran, NYU: Terminal value — stable-growth terminal claims.
- TradingView: Bar states (paid link) — confirmed-bar behavior used by the indicator.
- TradingView: Pine Script v6 reference (paid link) — language/API reference for compilation checks.
Final Takeaway
Borrow research questions from successful investors, then own the reasoning. A usable disclosure process reconciles dates, claims, units, weights, valuation and cash. A thoughtful decision to defer an idea is a successful workbook outcome when the evidence is incomplete.
Risk Disclosure
This material is educational and is not personalized investment, legal or tax advice. Stocks, derivatives, short positions and concentrated portfolios can cause substantial losses. Historical examples and hypothetical calculations do not predict future results. Prices can gap, liquidity can disappear and estimates can be wrong. The indicator is an uncompiled educational observation tool and does not simulate execution or establish profitability.
This independent companion is not affiliated with or endorsed by John Konnayil Vincent, Wiley, the profiled managers, the SEC, Amazon or TradingView.