Situational Awareness should've built this risk management dashboard

​

You've probably seen by now Situational Awareness LP's public book got smoked last week.

This has been a fund I've been writing about since beginning of this year.

Funny enough, you'd actually be up ~41% since we first covered how to copytrade his US positions back in March 1st.

Both can be true...

You can be right on the stocks and still be forced to sell.

From the public evidence, two vulnerabilities appear to have mattered:

1. Leverage shortened the amount of time the fund could afford to be wrong.
2. Liquidity and crowding made it harder to change the book once the trade reversed.

As Warren Buffett once joked - smart people tend to go broke through liquor, ladies, and leverage. The real warning was leverage.

I would add a fourth L: liquidity.

Everyone on X is talking about SALP's leverage issues but fewer mention the liquidity on the underlying names they held.

Trading crypto in college in 2017, I went through a roughly 90% drawdown by holding too many low volume names. In hindsight, it was cheap tuition before I managed money institutionally.

Piling into lower volume assets (especially with leverage or shorting) makes you vulnerable to sharks attacking your positions.

In Situational Awareness's case here, the shark was Citadel.

So today I'm going to show you a risk management technique using Codex to build this liquidity risk dashboard.

It answers two simple questions:

1. How many trading days would it take to exit each position?
2. Which positions appear across other Form 13F filings, creating potential common-exit risk?

This risk management technique is especially relevant if you trade lower market cap companies (e.g., biotech, small-caps) or for your short book, and one I run for these sorts of situations.

How to Build This Dashboard

I asked Codex one question:

If I had to liquidate this portfolio, which positions would trap me?

Here is the high-level prompt:

Full prompt at bottom of dashboard: Link​

The Result

Full output here: Link​

Three things jumped out.

(1) Roughly 1/4 of the disclosed stock book was stuck behind a 2 week exit window

These were not positions you could unload in an afternoon without becoming the market.

Which makes you a prime target for engineering squeezes.

(2) The worst bottleneck would take months, not days, to unwind.

$CORZ for example would've taken over 11 weeks to unwind! You become the market at that point and very vulnerable. Many funds I've heard were shorting once they smelled blood.

(3) Several of the hardest positions to exit were also popular with similar funds.

That matters because average trading volume looks like liquidity only when you’re the only seller.

During a forced unwind, everyone is trying to fit through the same door.

This is how the risks compound:

• Leverage starts the clock.
• Liquidity determines how quickly you can move.
• Crowding determines how many people are trying to leave with you.

Six of the seven hardest disclosed exits also appeared in at least five similar portfolios.

Most risk dashboards tell you how much a position could lose.

This one tells you whether you will still control the decision to sell.

A good risk system should warn you before the P&L does!

And thankfully AI can help you vibecode these dashboards quickly.

How you can wire this up

I included the prompt I used in the bottom of the output link above.

If you want to take this to the next level and use this as an active monitor, I would connect it directly to:

• My live portfolio and actual position sizes
• A market-data API tracking volume, volatility, spreads, and borrow
• Your leverage, financing, and risk limits
• Slack through a webhook or MCP

Then I would have it recalculate continuously and trigger an alert whenever liquidity deteriorates, an exit window crosses my limit, or the time required to unwind becomes longer than my financing runway.

The Slack alert could be as simple as:

Now you have an automated early warning system!