Superforecasting for Dummies
Dolla Dolla Bill, Y'all!
Is predicting the future hard? A lot of people say it is, but I really don’t think so. To prove it, I decided to do a little experiment. In the past, whenever I told people that I had refined superforecasting principles into a real science which was superior to economics, I was always met with variations of the same skeptical comment. “Oh yeah, Mr. Big Brain? Well, if superforecasting is real and you can predict the future, why haven’t you used that ability to make a lot of money on the stock market?” Eventually, I got tired of these sarcastic responses, so I decided that the easiest way to prove superforecasting is real would be to just follow their suggestion and make a lot of money on the stock market.
The problem with doing something thought to be impossible is that people are so convinced you can’t do it that they will try to come up with some excuse - any excuse - to debunk you even after you have already accomplished the “impossible” task, because acknowledging that you have more knowledge than the “experts” would force them to question their underlying assumptions about the way the world works. This would be both scary and humbling for them, and in general people don’t enjoy being scared or humbled. In other words, before you do something impossible, it’s important to make sure to document it thoroughly so that your haters can’t find an excuse to diminish your accomplishments. Even then, understand that many of them will refuse to believe you anyway regardless of how thoroughly you document your evidence. Your goal should not be to try to convince everybody of the superiority of your theorems, but simply to convince enough people for you to have an overwhelming tactical advantage over your haters. That way, if push comes to shove and they try to discredit or deplatform you, you’ll have enough people on your side that you can push back harder and inflict far more damage to them than they could possibly inflict on you.
The first step in proving the validity of my superforecasting science was to quantify my target goal. One truism among investors is that the average investment doubles every seven years. I decided that a good way to prove the superiority of superforecasting over classical economics was to triple my investments instead of doubling them. And instead of tripling them in seven years, I was going to triple them in three years.
At first glance, that may not sound so impressive. It’s only once you calculate compound interest that the true potential of this investment technique becomes obvious. Let’s compare how my rate of return would compare to the standard rate of return over a 21 year period. Assume that the normal investor and me both start with $100.
By year 7, the normal investor has doubled his investment to $200. By year 14, he has doubled his investment again, to $400. By year 21, it has doubled one final time, to $800. Now let’s compare that to my rate of return.
By year 3, I have tripled my investment to $300. By year 6, I have tripled it again to $900. By year 9, I have tripled it again to $2700. By year 12, I have tripled it again to $8100. By year 15, the investment is worth $24300. By year 18, the investment triples to $72900, and by year 21, it has tripled one final time to $218700. In other words, over a 21 year period, my investment strategy would make 273 times more money than the standard investment strategy! Those are insane numbers, and it’s easy to understand why somebody would be skeptical that I could achieve those goals. But data doesn’t lie. Here’s a graph of my investment returns over the past two and a quarter years. As you can see, I still have eight and a half months to go before hitting my target date of three years, and yet I’ve almost hit my goal already. In other words, not only are my investments beating the return rates of every single fund available to a retail investor, but I’m going to hit my goal much earlier than expected. That’s the power of superforecasting.
The second step in proving the effectiveness of superforecasting was to demonstrate intent. Anybody can have a lucky year in the market. Maybe you invested in Tesla at just the right time. Maybe you got into a meme stock like GameStop just before it blew up. It’s very easy to get lucky in the market and then retroactively come up with a good narrative like superforecasting to explain why you did so well, even though the real reason was just luck. I knew that my critics would immediately try to dismiss my superforecasting ability as a lucky streak of this sort, so I needed some sort of proof to demonstrate that my sudden run of “good luck” in the market was in fact premeditated.
I decided that the easiest way to prove my market success was the result of superforecasting rather than luck would be to write an entire book about superforecasting and publish it in 2019, before making the majority of my money on the stock market. If somebody accused me of just getting lucky in the market and then coming up with a convincing explanation after the fact, they would have to explain why I had published a book about superforecasting a few months before my sudden “lucky streak.” That would have to be a really wild stroke of coincidence, wouldn’t it? As a former programmer (and music video aficionado), I like leaving easter eggs hidden inside my artwork, so I also made sure to drop a lot of references to viruses in Chapter One to subtly allude to the pending Covid pandemic that we would experience the following year. If you examine the date that my book was published and compare it to the graph of my investment returns, you’ll see it came out in October of 2019, before Covid hit and also before I made the majority of my money. This is all starting to get really hard to explain away by “coincidence” or “luck,” don’t you think?
The third step in proving the validity of superforecasting is to establish credibility. After all, to you I’m just a random stranger on the internet. Anybody can make up a fake graph and lie about their stock returns. So, I’m issuing an open challenge to anybody who questions my credibility. I’m willing to give up the safety of my anonymity in order to verify my claims. Any venture capitalist or hedge fund director who doubts these numbers is welcome to tell me they disbelieve me and that they want me to prove my data is accurate. For the price of a round trip airline ticket, I am happy to fly down to wherever you are and authenticate this data beyond the shadow of a doubt for you. If I’m proven to be lying, I will gladly append a retraction to the end of this post and refund you for the cost of my plane ticket. However, understand that I would be risking a lot of blowback by revealing my identity, since many people might get unhappy about the way I manipulated the 2016 presidential election. So if I can demonstrate to you that my returns are accurate and that I haven’t fudged the data in some manner, you’ll have to compensate me by offering me a job as a portfolio manager (if you run a hedge fund) or help me secure venture capital funding so that I can set up my own investment company (if you are a venture capitalist). I hope that this open challenge to anybody who disbelieves me can demonstrate the legitimacy of my superforecasting science.
The fact is that the future is quite predictable through superforecasting, but economic “experts” don’t want to admit it since it would demonstrate that much of their own field is a lie - that they are essentially the equivalent of medieval alchemists trying to compete with modern chemistry. They have too much status and reputation invested in the current status quo to ever accept being humbled by anybody who lacks the long string of academic credentials currently required to claim expertise in their field. However, I don’t recall asking for their approval. As I described in one of my earlier posts, this behavior is not new. Historically speaking, primitive scientists have often refused to accept the modernization of their paradigms because they have no skin in the game. The only way to convince them to accept any new scientific paradigm which would reduce their own status is to weaponize it so that they don’t have the option of denying the new consensus reality. Fortunately, the science of predicting the future is extremely easy to weaponize, and in my next couple of posts, I’ll share a few ideas on how to do that.