- The Investors Collaborative Newsletter
- Posts
- What happens when you focus on win rate?
What happens when you focus on win rate?
Part 1: Overfitting is fun sometimes, right?
So, let me preface this with “this is all experimental and shouldn’t be trusted just yet!”
This is an experiment, a giant test to see what happens. I want to see what we can learn from this entire process, not to come up with a highly usable strategy right off the bat. As always, read it, understand the logic, and assess the risks involved and see if they’re to your liking. If you can fit some science into your portfolio without making yourself freak out? Go for it!
Additionally, if you’re following along with this, don’t have access to the WHSMacon database, and would like access to it, contribute strategies to the community! We don’t know how many, or if there are any additional requirements, but from the riddles we’ve seen posted its very clearly gated by contribution of strategies to it. Not confident in your strategy? Doesn’t matter, no one is going to judge you (unless we think you’re trolling). Think it’s boring? Good! Boring strategies that work are the best kind of strategies. Is it an exact copy of another strategy and you’re trying to cheese your way to access? That’ll probably get you banned whenever the operator figures you out, don’t do that.
All of that understood? Good!
But now, a word from this weeks sponsor!
Does looking at your portfolio make you want to drink? Well, I’ve got 2 solutions for you. Run better strategies, and check out this weeks sponsor!
Whiskey: A Hedge Against Market Volatility
Looking to protect your portfolio from the next recession?
Consider investing in rare spirits like whiskey.
Whiskey investing provides a proven hedge against stock market dips driven by inflation and other factors.
With Vinovest, you can invest in high-growth segments such as American Single Malt, emerging Scotch, Bourbon, and Irish whiskey. Thanks to established industry relationships, Vinovest overcomes industry barriers that have made historically whiskey investing expensive and opaque. As a result, you can enjoy high-quality inventory that boosts your portfolio value and enhances liquidity.
Now, back to the experiment
Step 1 is trying to figure out how to screen the db for strategies. First, lets establish goals.
Goal 1. Positive win rate.
No one likes seeing red days. Lets Try and make sure it’s at least 50/50
Goal 2. Makes more when they win than when they lose.
Self explanatory, we want strategies that, you know, work.
Goal 3. No cash positions.
I don’t want to be sitting out of the market. Ever. Sitting out means no gains, and we’re not looking to sit out coin flips, we’re looking to guess them correctly. Ironically, I don’t view UUP as a cash position, it just moves too much for that.
Goal 4. The ending win rate is higher than the average of the constituents
Whats the point of running multiple strategies if it doesn’t improve the statistics?
Now, the specific criteria to screen for;
The following tickers are not present;
SHV
SHY
BOXX
CSHI
ICSH
GSY
FLOT
BSV
BIL
Win Days > 51%
CAGR ≤ 1000%
OOS on or before 3/21/2024
Start Period on or before 1/1/2024
Payoff Ratio ≥ 1.25
Profit Factor ≥ 1
That long list of tickers is there to make sure the win rate isn’t being cheesed (artificially boosted for the non-gamers among us) by things that just sit in BIL, BSV, or your short term bond ETF of choice. We’re looking for logic that actually uses risk on tickers. Now those risk on tickers could be something as simple as GLD, UUP, or SPLV, but movement is movement. Movement means it can be going up, and going up is the end goal.
Additionally, I want to see things with a positive win rate and positive expectancy plus a little extra. Additionally, I want to see things at least 6 months old stuff, and things that will backtest to at least the covid crash. Could it be longer? Yes, absolutely, but this is part of the testing and theorycrafting. I need some kind of minimum standard on length, and any longer would cut out so many strategies.
Now, this gives us a list of 30 strategies. I do not like all 30 strategies, and I don’t like the chart of a bunch of them. I also do not like the size of a bunch of strategies, that’ll make later stages of this process much more difficult for no real benefit. So lets narrow this down to ~10.
The 10 strategies
Now, as you explore these strategies, you may start to notice something.
All gas, no brakes
These are not at all within the real of things I’d normally run. Well, well outside of my tolerances and desires in a strategy. Lots of volatility, questionable hedging, terrifying levels of concentration in volatility assets, and likely a healthy dose of overfitness. I have not dug in to the logic, cleared up any “funky” asset allocations, or done any testing on the sub blocks. Hell, when this process is finished, I would not be surprised to see half these strategies booted from the mix for varying levels of not passing any of the test I plan on doing. Without further ado, I present to you the equal weighted, combined grouping of the above strategies. If you decide to run this, I have no sympathy for you in any way. You do so at your own peril. Its a science experiment, nothing more, nothing less.
Interestingly, in the Maestro charts, you can see the strategies are, somehow, miraculously uncorrelated even during periods of drawdown. That said, I would like to add in something that is as uncorrelated as possible, but I don’t know if that’ll be feasible with the screening limitations I’ve put on the strategy pool.
So, whats next?
The next couple steps, and the next couple issues, will be going through some of the following steps.
Clear out underperforming, redundant, overly concentrated, and non-fractionable assets.
Verify the logic all serves an actual purpose.
Condense redundant logic. The same frontrunner doesn’t need to be present 4 times in 4 separate strategies, we can just condense all the endpoints under one frontrunner.
Turn this pile of strategies into a portfolio I’d actually use
And that’s going to wrap it up for this week!
A bit of a setup for a longer series, coving a large swathe of the strategy creation and/or cleanup process. Keep your eyes peeled for future editions where I’m going to detail how to check the win rates of individual assets, whether the stats support inclusion of an asset/check, and other “quality control” measures for your strategies you’ll likely find use in employing!
Housekeeping time!
Whiskey: A Hedge Against Market Volatility
Looking to protect your portfolio from the next recession?
Consider investing in rare spirits like whiskey.
Whiskey investing provides a proven hedge against stock market dips driven by inflation and other factors.
With Vinovest, you can invest in high-growth segments such as American Single Malt, emerging Scotch, Bourbon, and Irish whiskey. Thanks to established industry relationships, Vinovest overcomes industry barriers that have made historically whiskey investing expensive and opaque. As a result, you can enjoy high-quality inventory that boosts your portfolio value and enhances liquidity.
Thanks again to our sponsor Vinovest this week, I’m a big fan of uncorrelated investments, and I can’t imagine the correlation is very high to more traditional assets.
Reply