Winning the Loser's Game — Chapter 2: The Catalogue of Unforced Errors

"The investor's chief problem — and even his worst enemy — is likely to be himself." — Graham, quoted repeatedly by Ellis

What This Chapter Does

Chapter 1 said outcomes are decided by errors. So the errors must be enumerated, or "make fewer mistakes" is merely a correct platitude.

They are ranked here by cost, and the ranking contains a counterintuitive finding: the most expensive error is not picking the wrong stock.

The Catalogue, Ranked by Cost

Rank	Error	Typical cost	Visible?
1	Selling during a decline	30%–50% of the portfolio in one act	Obvious in hindsight
2	Paying high fees for decades	15%–25% of the 30-year ending value	Nearly invisible
3	Over-concentration	Potentially everything	Obvious in hindsight
4	Chasing recent performance	2%–5% each time, repeatedly	Invisible
5	Frequent trading	Friction plus taxes	Partly visible
6	Allocation mismatched to horizon	Varies	Invisible

The difference between ranks 1 and 2 is not magnitude but visibility. You know what you are doing when you commit the first. The second happens a little every day for thirty years and is never noticed.

Rank One: Selling During a Decline

The only error capable of destroying an entire retirement plan in a single act.

Its mechanism was proved in our Retirement Decumulation Mechanics Chapter 2: shares sold do not participate in the recovery. And someone who exits during a decline almost always returns after the rebound — because the fear that drove him out only subsides once recovery is evident, by which time prices have already risen.

Why it is "unforced": nobody compels you to sell. A falling market demands nothing of you. This error is entirely self-generated — precisely Ramo's meaning.

The only effective defense is structural, not volitional: the cash buffer of Decumulation Chapter 3, the written policy statement of the Defensive Investor's Operating Manual Chapter 1. Relying on "I will stay calm next time" fails, because you are not the same person during a panic.

Rank Two: Fees — The Expensive Invisible Error

$100,000 over 30 years at 7% nominal:

Annual fee	Ending value	Lost vs. 0.03%
0.03%	~$754,000	—
0.50%	~$660,000	~$94,000
1.00%	~$574,000	~$180,000

1% sounds like a small number, which is exactly why it works. Nobody would agree to pay 1.8× their starting capital in fees — yet charged as 1% across thirty years, almost nobody refuses.

This is the same fact as Random Walk Chapter 5's cost argument, from a different angle: there it is "cost is the most reliable negative predictor" (statistical). Here it is "this is an error you can permanently eliminate today" (operational).

Rank Three: Over-Concentration

The only error that can produce permanent, unrecoverable loss.

Other errors reduce your returns; concentration can take you to zero. A company can fail and an industry can vanish; a diversified market does not.

Common forms:

• Employer stock. Your salary and your investments ride on one company — if it fails you lose both simultaneously.
• Inherited or long-held positions with large embedded gains, untouched for tax reasons (the trap named in Defensive Investor's Operating Manual Chapter 3).
• "I understand this company." Familiarity is not a substitute for diversification; it only makes concentration feel safer.

Rank Four: Chasing Recent Performance

This error's signature is that it looks reasonable every single time.

You sell the fund that lagged for three years and buy the one that led. Each step feels like deciding from data, and the aggregate effect is systematically buying high and selling low.

The persistence data in Random Walk Chapter 5 is the direct rebuttal: a fund in the top quartile over five years has close to random odds of repeating. Three- and five-year records carry almost no predictive information — and they are exactly what fund marketing emphasizes.

A Boundary That Must Be Clear

"Fewer errors" does not mean "do nothing." This is the book's most easily misused idea.

These are not errors but required actions:

• Rebalancing by rule, even when it means buying into a decline
• Withdrawing living expenses on schedule
• Adjusting allocation for a genuine change in life stage
• Trimming an over-concentrated position

The test is simple: an action driven by a rule written in advance is not an error. An action driven by today's emotion or headline is.

Procedure

1. Audit the last five years against the six items and mark honestly which you committed.
2. Fix rank 2 first, because it is the only one eliminable permanently today. Check existing holdings' costs with /tools/portfolio-simplifier.
3. Build a structural defense against rank 1, not a promise: cash buffer plus a written policy statement.
4. Check concentration: any single holding above 5%, or employer stock above 10%, needs a reduction plan.
5. Stop replacing funds on three- and five-year records. Put that in the annual checklist.
6. Repeat this audit annually. The catalogue's value is in reuse, not in having read it once.

Relevance to a Retirement Portfolio

For retirees the ranking shifts — rank 1 becomes considerably more expensive.

The reason is sequence risk (Decumulation Chapter 2): an accumulator who sells during a decline loses compounding time; a retiree who does the same simultaneously locks in the loss and permanently reduces the base for all future withdrawals.

Rank 6 also changes shape in retirement. During accumulation, a horizon mismatch usually means being too conservative. In retirement it can err in both directions — too aggressive meets sequence risk, too conservative fails to outrun thirty years of inflation. Assumption 2 in Decumulation Chapter 1 makes this point: over-conservative portfolios fail more often.

The chapter's final meaning: retirement success is a subtractive process, not an additive one. You do not need to find a single brilliant investment. You need to not do the six things on this list.