“Leverage is the only way a smart person can go broke.” – Warren Buffett
Leopold Aschenbrenner might be the smartest guy in the room. He was 22 when he was fired from OpenAI, then wrote “Situational Awareness,” a widely-read essay predicting the arrival of Artificial General Intelligence (AGI).
Aschenbrenner turned that essay into a hedge fund, Situational Awareness. The pitch was simple: go long AI infrastructure (chipmakers), bet against the legacy software names getting left behind.
Chips up, software down. For over a year, it was the trade of the decade. The fund was up 439% in the first half of 2026, ballooning to $45 billion in assets.
Aschenbrenner was so confident in his investment thesis he borrowed money to juice returns. He ran the portfolio at roughly 4x leverage for every dollar invested. Debt amplifies investment outcomes; both good and bad. When chipmaker stocks (semiconductors) violently corrected in July and the short side went against him, the margin calls came fast. The fund lost 67% in a matter of weeks. He avoided full liquidation on July 30th by fire-selling $10 billion of stock to Citadel.
He was right about the direction of the AI trade. He wasn’t right about how much to own.
Here’s the lesson investors should heed; his thesis wasn’t wrong. His position size was. Leverage doesn’t just amplify your returns — it decides whether you’re still at the table when your thesis finally plays out.
It turns out there wasn’t much situational awareness at Situational Awareness (bad joke).
Most retirees aren’t running 4x leverage on AI chip stocks. But a lot of them are getting position sizing wrong, and they don’t know it.
Stealth Concentration
Own the S&P 500 through SPY. Own the Nasdaq 100 through QQQ. Own the Magnificent 7 outright (Amazon, Microsoft, Meta, etc.).
It feels like diversification. It’s the equivalent of going to In & Out Burger and ordering a bacon cheeseburger and a hamburger add the cheese & bacon. It’s the same order said in a different way.
The Magnificent Seven represents ~35% of the S&P 500. The S&P 500 and Nasdaq essentially share the same top eight holdings. An investor is not buying two different risk & return profiles. They are buying the same bet twice, with different ticker symbols.
SPY (S&P 500 ETF) Top Holdings
QQQ (Nasdaq ETF) Top Holdings
The above graphics show the top eight holdings for the S&P 500 (SPY) & Nasdaq Index (QQQ). The percentage weighting of each stock might be slightly different, but it’s virtually the same exposure.
It gets worse; most investors own a direct position in the same stocks outright. The result is a triple bet on the same volatile theme, AI & Technology.
This usually doesn’t happen intentionally. For one, most people don’t know how much redundancy there is among popular stock indices. Throw in multiple accounts spread across various financial institutions, financial advisors, old 401k plans, etc. and overlap is going to happen.
When we aggregate exposure for prospective clients across every account, we often find 40-50%+ of total household equity riding on the same eight to ten companies. Nobody sat down and decided to make that bet. It accumulated, one reasonable-sounding decision at a time. That’s what makes it stealth: no single account is the culprit, but the aggregate concentration creates a massive, unintended bet.
Nobody notices concentration risk on the way up. It’s been said a bull market hides a ton of warts. Everybody notices it on the way down, and by then it’s too late to do anything.
How can we uncover unintended, concentrated bets before it’s too late?
We’ve got a portfolio X-ray tool that unveils exactly what an investor owns across accounts, advisors, and financial institutions…
Source: Koyfin (8/5/2026)
The above chart shows what percent of a sample portfolio that is held in each stock. The software can scan ETFs, mutual funds, and separately managed accounts to itemize how many instances the stock is owned. In the above example, NVIDIA makes up 3.30% of the portfolio and is owned in 9 different funds. Most investors would never identify the redundancy by looking at their statement.
In virtually every instance we run this exercise with prospective clients, we uncover a massive active bet the investor had little intention of making. In investing, it’s fine to take known risks, but it can be perilous to take unknown risks.
Position Sizing for Retirees
If you’re retired and drawing income from your portfolio, the goal is to stay in the game. It’s no longer “how do I make the number bigger.” It’s “how do I make sure the number is still there in twenty years, generating the income I need.”
This reframe should drive every position sizing decision:
- No leverage. None. A retiree living off portfolio withdrawals has no business borrowing to invest — the math only works if you never need the money at the wrong moment, and most retirees need the money to supplement their lifestyle.
- Be mindful of single-name exposure across every account you own — not just the ones you can see individually. If NVIDIA is 3% of your brokerage account, 8% of your S&P 500 index fund, and sitting inside your QQQ ETF, your real NVIDIA exposure isn’t 3%. Know what you own before a market event happens.
- Ask the asymmetry question before every large, active bet: if this works, does my life materially improve? If it doesn’t, you’re taking retirement-plan risk for a reward that doesn’t change your life. That’s a foolish bet to make.
A retiree would do well to size their portfolio positions so that no single outcome — chips down, software up, AI correction, whatever it turns out to be — can undo forty years of investing, saving, and building. We like to say positioning beats predicting.
For most retirees, the goal was never the relentless pursuit of more regardless of risk. The goal is to do what you want, when you want, with who you want with little concern for a portfolio implosion.
For more information on position sizing and prudent risk management, shoot us a note at shoot us a note at [email protected].
