The Statistics Are About the Index. The Pitch Is About a Stock.
The most dangerous number in a pitch isn't the fabricated one — it's the true one about something else. One question catches it: what population is this statistic about, and is my deal in it?
This week a newsletter writer published a piece arguing that it's better to buy stocks at record highs than to wait for the next crash. And here's the thing — the first half of the article is right.
The data is real and it's been studied to death. Going back to 1926, the S&P 500 has set a new all-time high roughly every six weeks. That's not a warning sign; it's just what an index that compounds for a century looks like. A market can't go up 100% without going up 20% first, and it can't do that without spending a lot of time at record levels. And buying at those records has historically been fine: a year after a new high, the index has been higher about four times out of five. If you sit out waiting for a 20% crash, the market can rally 30% first, and your "cautious" entry ends up more expensive than the record high you were scared of.
All true. If the article stopped there, it would be a solid case for boring index investing, and I'd have nothing to write about.
It doesn't stop there. Watch where it goes.
The bait and the switch
After establishing the buy-at-highs data, the article pivots. The author says he's adding risk now, partly because a chart signal he follows — a specific pattern of institutional buying — flashed positive. This comes a few paragraphs after he correctly explains that nobody can time the market. Then the destination comes into view: he's buying individual stocks down the AI supply chain, he has a new highest-conviction position, and he'll reveal the name — along with five more picks — in his paid newsletter.
Read that arc again. The evidence is about the index. The sale is about stock picks.
Every statistic in the first half of the article describes the S&P 500 — a diversified basket of five hundred companies that automatically deletes its failures and replaces them with winners. The index can wait out any drawdown, forever, because it isn't any particular company. That's precisely why buying it at a high keeps working.
None of that transfers to a single stock. Ask anyone who bought Cisco at its record high in March 2000 — the most valuable company on earth at the time, the picks-and-shovels play of that buildout. The index went on to set hundreds of new highs. Cisco took more than two decades to see its old price again. Buying "the market" at a high and buying "a stock" at a high are completely different bets. This article dresses the second one in the first one's statistics.
I don't think that's an accident, and I don't think it's unusual. I think it's a template. Which is why LPs should care.
Deal decks borrow statistics the same way
The same move shows up in private placements constantly, and it's easy to miss because the statistics quoted are usually true.
"Multifamily has never lost money over a ten-year hold." "Real estate always recovers." "Private credit has outperformed through every cycle." Versions of these lines appear in half the decks that hit my inbox, and as claims about the asset class, they're roughly accurate.
But I'm not being offered the asset class. I'm being offered one deal.
The asset class doesn't have a loan coming due in 2027. My deal does. The asset class doesn't have a preferred-equity slice that gets paid before I do, or a fee stack, or a sponsor who can run out of money. My deal has all of those. The asset class can wait out a bad market indefinitely, exactly like the S&P 500 can. My deal has a clock on it, and the clock is what kills deals.
The 2021 vintage proved this precisely. On a ten-year view, multifamily as a category was and is fine — the ten-year statistic held. The deals died anyway, in year two and year three, of floating-rate debt maturities the category-level statistic knows nothing about. The number in the deck wasn't false. It was about something else.
The question that catches it
So here's the check I run whenever a deck leads with a statistic: what population is this number actually about — and is my deal in that population?
It usually takes one honest minute. A diversified index versus one company. An unlevered asset class versus my position behind this specific debt and these specific fees. "Real estate over any ten-year period" versus a deal that has to refinance in year three. Track records of surviving funds versus everyone who started. Each pair sounds like the same thing. Each pair is a different bet.
| The statistic's population | What I'm actually buying | What the population survives that my deal doesn't |
|---|---|---|
| The S&P 500 at a record high | One stock at a record high | Any single bankruptcy — the index deletes its failures; Cisco took two decades to see March 2000 again |
| Multifamily over any ten-year hold | One deal with a loan coming due in 2027 | A drawdown of any length — the category has no maturity date, my deal has a clock |
| An unlevered asset class | My position behind this debt, this pref, and these fees | A loss that wipes out the equity but barely moves the category average |
| Track records of surviving funds | A fund that still has to survive | The failures — they left the sample the statistic was computed from |
The simplest version of the question: could this statistic survive something that would kill my deal? The S&P 500 survives any single bankruptcy — that's the whole design. The asset class survives any single foreclosure. If the number being quoted belongs to a group that can't go broke, and I'm being asked to fund a thing that can, the number isn't evidence. It's decoration.
The true number about something else
I want to be clear that I'm not calling anyone a liar, in the article or in the decks. That's what makes this pattern worth writing about. The most dangerous number in a pitch isn't the fabricated one — fabricated numbers get caught. It's the true number about something else. It survives every fact-check, because it's accurate. It just isn't about the thing you're buying.
This is the third time I've written about the same instrument lately. First it was the language — the late-cycle phrases that show up near tops. Then it was the story — how a thesis can come completely true while the investors funding it lose everything. This one is the statistics. Different tells, same discipline: the pitch is always about the category, the future, the pattern. Your money is always in one specific deal. Keep asking which one the evidence describes.
Related: The Euphoria Scorecard and Civilization Got the Returns. The Investors Got the Bankruptcies.
See also: A Lower Price Isn't a Discount and The Bear Case for Stocks Is a Sales Pitch in Your Inbox
This is one of many checks I run on every deal I consider — not the whole diligence process. It isn't investment advice, and nothing here is a recommendation to buy or sell anything. Do your own homework — that's kind of the whole point.
Tim is an LP in real estate and alternative credit and builds DiligenceBrief, a due diligence tool for LPs.