OK, so now that I have the warnings out of the way and a guide to proper back-testing established, we can talk about the actual method for developing your own systems. It all starts with a core theory. A core theory could be something as simple as NBA home underdogs cover the spread more often than not. This theory could be completely wrong, but it’s a starting base for data exploration. The core theory establishes a baseline result and represents the first two rules in the system (home team, underdog).
In this example, let’s say NBA home underdogs cover the spread 57% of the time. This is a good starting point in theory because it already has a high enough win rate to overcome the vigorish (that’s the cut the house keeps), however, this isn’t even close to what is needed to make it a ScoreMetrics system. This is when we get to the tricky part. We need to add layers of rules to improve the historical performance. The problem lies in where we begin to find these additional rules and how to decide whether it should be included in the system.
The ScoreMetrics lab is where this magic happens. We start by looking at dozens of possible rules. For example, what happens to the performance if we only include games in which the home underdog is on a 2, 3 or 4 game losing streak. How does this affect the ROI of the system? Then we go to the next possible rule, and the next, and so on until we have evaluated a large pool of potential rules that can be applied to the system. The goal is to intelligently add rules to the system structure until we get to 3 to 6 rules, with a sweet spot of 4 to 5 rules.
By layering these rules as discussed above a lot is accomplished. First, we are narrowing the scope of the games in the system. A system that had hundreds of games in a season that trigger a trade is going to struggle quite a bit to perform well. Second, the rules add complexity, which, in turn, makes it pretty much impossible for anyone with knowledge of the system structure to be aware of how you are generating the alerts. Basically, we get to keep this a secret. Third, we are adding layers of rules designed to incrementally increase the performance of the system. The first two rules in the example generated a 57% win rate. Adding the 3rd rule raised it to 59%, the 4th rule raised it to 64% and the 5th rule brought it all the way to 68% win rate. When we reach the right number of trades in a season, the right amount of rule complexity, and the ideal performance, then we have a system we can analyze and judge for inclusion in the ScoreMetrics portfolio.
How to Use Multiple Systems to Diversify Risk and Get Optimized Performance
One of the key elements of the ScoreMetrics Method is a blend of multiple systems to create a portfolio that reduces individual system risk. To be clear, these systems are not correlated with each other. One system can underperform and it should have no effect on the performance of the other systems as they use different underlying logic. However, because there is no relationship, then the portfolio itself is not connected system by system, and thus it becomes a random portfolio of underlying trades.
This randomness reduces the overall risk of the system, as well as increases the baseline performance expectations. If, for example, one system underperforms and you were only betting that system, then you would be exposed for the full drawdown of the system. However, because you are combining that system with, say, five other systems, that drawdown is both a smaller percentage of your overall investment and mitigated by the average performance of the remaining five systems.
When I use the Score that I discussed in an earlier section, I calculate the per system drawdown risk. If a system has a per trade allocation of 3 units, and a max drawdown of 24 units before the system is removed from the portfolio, then I can effectively take a net 8 trade loss before stopping myself out of the system. This means I have to allocate 24 total units to the system as a worst case scenario. If we are using 6 systems for a particular sport season, and let’s just say for ease of discussion’s sake that all 6 systems used the same 24 unit max drawdown (they will likely vary in real life), then there is a total of 144 units. However, the ScoreMetrics Method operates based on a starting portfolio of 100 units. This is to say if you have $10,000 in an account, then each unit would represent $100. If you have a trade that is a 3 unit allocation trade then you would risk $300 on that trade (3 units X $100 per unit).
So how can you have a total risk of 144 units but only fund the account with 100 units? You can leverage theoretical margin to take advantage of diversified risk of the portfolio of systems. This is akin to a margined stock trading account. Stock trading accounts can be margin-enabled to allow you to take 50% of your stock portfolio and access it for additional funds. A $10,000 stock account effectively becomes a $15,000 account in that example. The belief is that, while the stock price will go up or down, the likelihood of the underlying stock going to 0 is very small, and if you have a portfolio of stocks then the risk of all the stocks going to 0 simultaneously is nearly impossible.
The same logic applies here by having 6 systems running simultaneously, all differentiated and not correlated with each other, you achieve a form of mutual fund-like risk diversification and therefore can trade the systems as if you have a 50% margin. This increases the aggressiveness of the portfolio, both from a potential risk and reward scenario. However, this is based on sound risk diversification logic and is a great way to get more exposure using the systems in a balanced and intelligent way.
If you are an experienced trader or investor at some point in this book you probably said to yourself “Ok. I get it.” If you are new to the trading game or looking at a first time investment, everything I just explained to you can be a bit daunting. Keep in mind that my team and I literally spent thousands of hours developing and testing these systems and years packing them all into a methodology that shows these types of results. I believe that if you read this material, digest it properly, and have the hours of time to dedicate to building your own systems, that it won’t just change the way you look at sports investing, but it will change the way you look at all of your investing.
I have said it before and it is worth repeating: ScoreMetrics changed my life for the better and I know it can change yours too. If you are interested in ScoreMetrics but don’t have the time to build your own systems, I invite you to join my VIP Members area and become one of the team for any upcoming sports season. I will warn you though; seats are limited and we do sell out each year.
If you purchased this book online, I will be sending out more information about upcoming seasons, our latest ROI’s, and new and exciting news that we find in our lab that could be useful to those of you that choose to build your own systems. I also hold free webinars a few times a year so keep looking for those updates in your email.
I wish you the best of luck in whatever you decide to do with your future investments, and I hope to hear from at least one of you that you are making more money and having more fun than me!
But most of all, I really hope you can get a piece of what I cherish the most. Freedom and the peace of mind of knowing that my investments are working and safe.
Until next time,
John J. Todora