I was sick for most of the past week, but I’m up and about now. The team, however, has been busy working on taking the AI algo that we have for trading BTC and adapting that for stock trading with MSTR and similar crypto stocks.

As the market appears to be turning around, I thought it might be helpful to explain how we validate new strategies. Many of the mechanics used here are likely to help you even if you are merely trading memes.

Let’s start with the baseline – our AI BTC algo (a version of this is used for part of our the hedge fund now, but you are witnessing data for a strategy that we don’t use – this is purely for educational purposes). The bright green is the BTC Alpha Fund and the orange is BTC’s performance – both since 2018.

You can see that the strategy already works quite well–it’s never had a down year. So, you want to start with something that works pretty well on its own–call that the original signal.

• A baseline signal could be: buy and hold MSTR when its price is above the SMA 200, sell to cash otherwise.

Now, we’ll begin adapting it to a stock holding strategy. Initially, this poses a difficulty: we won’t be able to short BTC on declines–because that isn’t permissible in many retirement accounts and we want to run this in regular retirement accounts.

The resulting problem is that we’re going to have long periods of “dead money.” We need to solve for that at some point.

As a bonus, however, we’ll be able to use MSTR rather than BTC, since MSTR is more volatile and promises better returns. There are two drawbacks here, though. A first is that MSTR wasn’t always a “Bitcoin” fund. It only really began to operate that way in 2020, so we’ll need to start with 2020 data. Next, because MSTR is more volatile in both directions, we’ll need to develop a more cautious strategy.

Below is a simple “copy and paste” of the AI BTC long signal into MSTR. The result is … fine. About what I’d expect. MSTR’s performance is in blue while the AI algo is in red. 

Overall, about equal performance (800% since 2020) but with much less volatility. This does its job in making intelligent risk-adjusted bets.

Still, it leaves a lot on the table. And while it’s less volatile, it’s not without significant 28%+ draw downs in a single week. Let’s start by improving some of those returns.

If you look at MSTR’s performance, you’ll notice massive spikes followed by quickly retreating prices. This is traded in the traditional stock market so it’s not difficult to understand what’s transpiring: a short squeeze.

To catch single day returns, simply introduce a profit taking strategy – something along the lines of the following:

• Adjustment 1: If your trade makes 20% or more on a single day, then sell 33% to cash and let the rest ride as normal.

Why 20%? 

Well, in TradFi that’s an amazing return for a single day. It’s 2x the annual return of the S&P 500. I find that a “gut feel” approach like this is better than statistical methods because it prevents you from overfitting your data set.

Next, let’s add an additional profit taker.

• Adjustment 2: For any given trade, if the profit exceeds 41%,then sell 33% to cash and let the rest ride as normal.

Why 41%? Gut feel. It’s a little over twice our “squeeze trade” concern. It’s good enough.

Finally, let’s add a “soft” stop loss. A hard stop loss is likely to inhibit returns with a volatile asset like this one and they often don’t execute as desired. Instead, we’ll use the following rule.

• Adjustment 3: If a given trade loses more than 10%, then sell 33% to cash and let the rest ride as normal.

Why 10%? Because that’s about 1 year of average performance for the S&P 500.

What do we get with these adjustments?

Much better! We’ve now boosted total returns about 200% and we’ve lowered volatility. This is a much smarter strategy.

We still have stretches of more than a year of dead money in this strategy. In the next newsletter, I’ll show you how we approach solving that problem. But for today, we’ve seen:

1. How to identify motives for profit taking (short squeeze action).

2. How to pick profit taking targets.

3. What a soft stop loss is.

4. How to pick a target for it.

Those simple steps, moreover, dramatically improved results.

Happy Trading!

                                                                               - Sebastian Purcell, PhD