Defensive Investor's Manual — Ch. 2: Choosing the Fund — A Four-Number Comparison
阅读中文版 (with Audio)"Buy a low-cost index fund" turned into a procedure: expense ratio, tracking difference, breadth, and structure — and why the cheapest headline fee is not always the cheapest fund.
🔊 Listen to Article (Chinese Audio)
The Defensive Investor's Operating Manual — Chapter 2: Choosing the Fund — A Four-Number Comparison
"For most investors, the best choice is a fund that owns the entire market." — the modern statement of Graham's conclusion
The Problem This Chapter Solves
The canonical Chapter 3 concludes: defensive investors should buy index funds. That sentence is no help at all when you open a brokerage page and find seventy funds with nearly identical names.
This chapter gives a four-number comparison. Its goal is not to find the "best" fund — a meaningless question — but to eliminate the clearly worse ones and then pick arbitrarily from what remains. For a defensive investor the selection process must itself be bounded, or it becomes another form of market timing.
Number One: Expense Ratio
The only figure you know in advance with certainty. Returns are uncertain; fees are not.
$100,000 over 30 years at 7% nominal:
| Annual fee | Value after 30 years | Consumed by fees |
|---|---|---|
| 0.03% | ~$754,000 | ~$8,000 |
| 0.20% | ~$718,000 | ~$44,000 |
| 0.50% | ~$660,000 | ~$102,000 |
| 1.00% | ~$574,000 | ~$188,000 |
The gap between 1% and 0.03% is $188,000 over thirty years — nearly twice the starting capital. That is not "slightly more expensive"; it is a different retirement.
Screen: broad index funds should cost under 0.10%. Mainstream total-market and S&P 500 funds run 0.015%–0.06%. There is no reason to select a broad index fund above 0.20%.
Number Two: Tracking Difference
The expense ratio is the sticker price; tracking difference is the actual bill.
Tracking difference = fund's annual return − index's annual return
A fund charging 0.03% that trails its index by 0.12% costs you 0.12%, not 0.03%. The gap comes from trading costs, rebalancing impact, and cash drag. Conversely, securities-lending revenue lets some funds trail by less than their stated fee.
How to check: find the "annual returns vs. benchmark" table on the fund's own page and read the last 3–5 years. Consistently close to the fee is good; erratic gaps indicate management problems.
Number Three: Breadth
| Type | Holdings | Coverage |
|---|---|---|
| S&P 500 | ~500 | US large cap |
| US total market | ~3,500–4,000 | Large, mid, and small |
| Global ex-US | ~8,000 | Developed + emerging |
| Global total market | ~9,000+ | Everything in one fund |
There is no single right answer, but there is one error to avoid: believing that more funds means more diversification. Holding an S&P 500 fund, a Nasdaq 100 fund, and a large-cap growth fund means owning the same companies three times — Apple and Microsoft appear in all three.
Test: list the top ten holdings of every fund you own and count the duplicates. More than five means you are using three funds to do one fund's job while paying two extra fees.
Number Four: Structure and Tax Location
The same index in a different wrapper produces a different after-tax result. The canonical never touches this layer because it is execution, not philosophy.
| Structure | Property | Best located in |
|---|---|---|
| ETF | In-kind creation minimizes capital-gain distributions | Taxable accounts |
| Index mutual fund | May distribute capital gains | Tax-deferred |
| High-dividend / bond funds | Generate continuous taxable income | Tax-deferred (IRA/401k) |
Principle: the more taxable income an asset produces, the more it belongs in a tax-deferred account. This is asset location, distinct from asset allocation — and most people do only the latter.
Procedure
- List candidates; keep only those under 0.10%. This usually cuts seventy to under ten.
- Check 3–5 years of tracking difference and drop any that trail materially more than their fee or do so erratically.
- Decide breadth: one total-market fund, or "US total + global ex-US." Never more than three equity funds.
- Run the top-ten overlap test. More than five duplicates means consolidate.
- Place by tax location: bonds and high-dividend assets into IRA/401k; broad equity ETFs into taxable.
- Record the choice in your policy statement and stop comparing. Agonizing between 0.03% and 0.04% converts the savings into anxiety.
Relevance to a Retirement Portfolio
Fees harm retirees twice, which deserves stating separately.
During accumulation, fees erode growth. During withdrawal they erode growth and principal simultaneously — drawing 4% while paying 1% means an effective 5% withdrawal, without any increase in what the portfolio can sustain. On a 4% withdrawal plan, a 1% fee cuts the safe withdrawal rate by roughly a quarter.
That is why this chapter opens with fees rather than returns. You cannot control the next thirty years of returns; you can control fees down to 0.03% today. Graham taught margin of safety — for a defensive investor the expense ratio is the cheapest and most certain margin available.