I think I've discovered a common cognitive error which I'm going to label the Physical Embodiment Fallacy.
In this and many other superhero movies, the bad guy creates/steals some horrifying new piece of technology which is going to kill millions, enslave the world, etc.
The solution in every movie is simple: kill the bad guy and destroy his evil lair.
I.e. eliminate the physical embodiment of the problem, and the problem is solved.
This is, of course, insane. The problem isn't the poison factory, it's the fact that someone has discovered how to create such an effective poison.
Even if you kill the guy, the knowledge doesn't die with him.
And even if, by some miracle, the bad guy was the only guy who knew the information necessary to create the superweapon, that's still not good enough.
All you need is the knowledge that a superweapon _of this type_ is even possible. That's more than enough.
You see this in trading all the time. When you're trying to find new trading strats, by far the hardest part is knowing where to look. What ideas to investigate.
Even knowing a tiny amount of information like "competitor X is making money doing Y" is HUGE.
Just knowing this is almost all you need. If you have smart people, an analog of Linus's Law applies.
"With enough eyeballs, all bugs are shallow."
If you know where to look, all problems are easy.
Genies never get put back into bottles.
Useful knowledge always expands to fill available brains precisely in proportion to its usefulness.
Bad guy super-weapons won't die just because the bad guy died.
/END
Addendum:
Maybe the essential thing that makes crypto people different is that they can see and avoid this cognitive error better than others.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
1/ The key to understanding "news" is to think about probabilities and incentives. A thread... 👉👉👉
2/ Let's say you're trying to estimate the chances of some future event happening.
As an example, let's take P(Trump runs for President in 2024).
3/ So you read the news about such an event. CNN, FoxNews, Politico, whatever. And it feels like a steady drip of "news" about tiny events which could affect that probability.
So how do we integrate all these pieces of information into a prediction?
1/ A short thread about the relationship between academic finance and real $ trading. 👉👉👉
2/ First of all, I need to get something off my chest. I loved the Asimov Foundation books as a kid.
Much like Paul Krugman, the idea of doing math to understand complex societies, and to predict their evolution, was transfixing. And still is.
3/ I used to do mathematical modeling of complex systems before I even knew it was a thing.
When I was working as an engineer (and running a lot), I spent more time that I'll admit trying to build a model of the human lactate response to exercise. For fun.
The better the trader, the less they care about which specific product/market they're trading.
That doesn't mean they don't have deep knowledge of the market or product. Far from it.
It means they don't *care*.
2/ Conversely, I see a lot of aspiring, new, and frankly bad traders who care a *lot* about the product.
"I trade options," "I trade futures" like it's a religious commitment. It's not. The product you're trading is a means to an end, at least if you care about money.
3/ One of the founders of my former company loved saying something like:
"If they made financial markets illegal tomorrow, we'd probably suffer for a while but we'd eventually be fine. We'll just go find something else to trade."
1/ A thread about the relationship between getting older and learning new things.
🧵👉👉
2/ It’s a weird relationship. One way of looking at it is through the lens of the explore-exploit tradeoff.
3/ In reinforcement learning, when you have to act in a novel environment and learn in an online way, there's a tension between trying new things vs doing the things you’ve already learned are good.
There’s a subtle but very real fallacy about backtesting that lots of smart quant-y people fall into. I’ve fallen into it many times. And arguably I still do, just in more and more subtle ways.
A thread 👉👉
1/n
So you have a trading strategy, and you want to backtest it to see if it’s any good. Being good boys and girls and others, we know we mustn’t overfit to the data we already have.
We know that historical data is precious gold, and it must be used carefully.
2/n
Well, imagine I propose the following solution: build a model of the market in all its gory detail: fat tails, heteroskedasticity, vol clustering, etc etc. I calibrate this model using historical data, and it’s pretty good.