# The Momentum Strategy's Invoice — July 2026, a -32.5% Month in a Live Account
- Source: FoldAlpha Research (https://app.foldalpha.com/en/research/momentum-crash-invoice) · Published: 2026-09-01 · Series: Myth Testing 10

Several articles in this series say that "momentum is a factor that survived." This article is
the record of paying that premium's price tag with a live account.

## July 2026

While running a momentum-based monthly strategy (Korea and US) in a live account, we took the July crash:

- Strategy account **-32.5%** (one month)
- KOSPI **-20.6%** in the same month
- About -12%p versus the market — when the market fell, the strategy fell more

This is not a coincidence. A momentum strategy structurally holds recently strong stocks = high-beta stocks.
Measured figures for one semiconductor stock in the portfolio at the time: on days the index rose +1%, it averaged +4.8% (upside capture ratio 477%);
on days the index fell -1%, it averaged -2.1% (downside capture ratio 210%); beta 2.79. This asymmetry is a blessing in a rising market,
and in a month like July it becomes the invoice, exactly. That stock went as far as -53% from its high.

## Momentum crashes are a documented phenomenon

The long-term performance curve of the momentum factor has deep crashes embedded in it periodically (1932 and 2009 are the famous ones).
This is precisely what academics point to as the reason the momentum premium has not disappeared for decades:
**because unbearable drawdowns arrive periodically**, institutional money that carries career risk
cannot hold on to this factor to the end, and so the premium remains.
In other words, momentum's excess return is not a free lunch but **the price of enduring the crashes**.

## What we changed and what we did not

- **What we changed**: we added a defensive rule that halves new allocations during overheated market phases following a crash. However, the threshold for this rule was set after seeing July, using that data (post-hoc calibration), so we noted in the documentation that pure out-of-sample validation only starts accumulating from now on
- **What we did not change**: the strategy itself. Overhauling a strategy right after a crash is the same as selling the premium at the worst possible moment, so we deferred the judgment to future numbers with a 6-month change freeze and pre-registered 18- and 36-month kill conditions

## What this record says

Looking at a momentum curve in a backtest, the crash section looks like a "valley you pass through."
In a live account, that valley is the experience of one third of your assets disappearing in a single month.
The question to ask before a strategy's expected return: **can you actually endure this strategy's worst month.**

## Data sources

- Realized P&L records from our own managed account (2026-07), Korea Exchange index data, self-collected US price data

---
This article documents tests on historical data for informational purposes only. It is not investment advice or a recommendation to buy or sell any security. Past test results do not guarantee future returns.