#220: Meltdown

micksonmatch
January 28, 2025

The temperature in Sydney is 40 degrees Celsius today, which is 104 degrees Fahrenheit. We are melting, and so were tech stocks yesterday.

S2N Spotlight

A few days ago a wise man described to me how a portfolio with a 10% allocation to Bitcoin would do versus the benchmark S&P 500. Let me share some charts and statistics, and then let’s talk about it.

The data goes back 10 years from today and shows a 90/10 portfolio rebalanced every quarter. The performance is amazing.

Let us dive deeper into the performance. The outperformance of the benchmark does look amazing.

Now let us look at some of the risk factors. A 63% drawdown is not a laughing matter, and to suffer 1109 days (3.5 years), I am not sure even Chuck Norris would have managed.

Look at the bold colours in the monthly table below; both up and down, those are massive numbers. Not sure this is a sleep-easy portfolio. More like, please book me an appointment with my cardiologist; I think I am having a heart attack….

This portfolio would be more in my wheelhouse, where I matched the portfolio to the benchmark’s volatility. This requires deeper explanation if you are not getting it. Today is not your lucky day, sorry (apologise to Clint Eastwood; I just had to).

Now for a dose of extra reality. It is easy to be wise when the S&P 500 benchmark and Bitcoin are at all-time highs. The question is how this portfolio would have looked 2 years ago after the FTX and crypto winter was in the depths of an existential crisis. One that NFT’s have not survived or barely, if you want to be generous.

Yes, the 90:10 portfolio still outperformed the benchmark on both an absolute and risk-adjusted basis. I am not sure most portfolio holders would have stuck it out though.

If you had asked me if I would have invested 10% of my portfolio in Bitcoin 10 years ago and stuck with it. The answer is a resounding no chance. Would I have said 1%? The answer is maybe. The problem is I would have been better off being 100% in the S&P 500.

S2N Observations

I think my observations about DeepSeek on Friday and Monday were spot on and will be an ongoing story over the coming weeks and months. An interesting story brewing about the founder of DeepSeek is that he apparently wrote the foreword to the Chinese version of the book about Jim Simons without the author Gregory Zuckerman’s knowledge or permission. He also gave a wildly embellished version of the secrets that were released in the book. Come to think of it, I don’t recall any secrets. I am starting to think that perhaps he is not the only owner of DeepSeek. Perhaps there are deeper pockets behind this project than we realise. Another interesting side note on this subject is that hackers are trying their best to hack the hell out of DeepSeek. I wonder who is behind this hack? I know I have such a suspicious mind, but it feels like state actors are all over this play.

If you were a buy-the-10%-dip kind of guy with Nvidia, this is how you would have performed after such an event. By the way, I didn’t try and make the threshold -16% as that would be curve fitting. Clearly any study that you apply to this stock over the last 20 years that went long would have made money. I would never invest in this stock based on this backtest strategy. Hindsight is such a beautiful view.

S2N Screener Alerts

Look at what we have here: Nvidia was down more than -16.5%, a massive 4-sigma down day.

Using normal distribution statistics, we get the following:

The probability of a 4-sigma down day for a stock is approximately 0.00315%.
This means such an event is expected to occur once every 31,746 trading days (assuming 252 trading days per year, this is roughly once every 126 years).