Statistics

You Can’t Predict The Unpredictable

There is, or might even not be, some BAD THING ahead of us. I don’t know what it is, and you don’t know what it is, and nobody knows what it is. But it could be BAD. It might be harmless, though. It might have something to do with money, or again it might not. Or it might not be anything at all.

Call this mysterious thing Y, for simplicity. What is the probability of Y?

What is the evidence for Y? That BAD THINGS happen? If Y is bad, and it might not be, then Y is at least possible. But good and indifferent things happen, too, and Y might be good or indifferent. So again Y is possible. And if it’s possible, then it has a probability, though maybe not one that’s quantifiable. What is this probability?

Too loose, too vague. The question doesn’t, and shouldn’t, make sense. So let’s change it so that Y is only BAD. How about if we call this mysterious Y a Fat-Tailed Black Swan? Now what is the probability of Y?

Y is a BAD THING. BAD THINGS happen! Everybody knows that. BAD is not good, BAD is something to worry about. We should do something. What about the children! Hurry, what is the probability of Y!

“What do you mean by ‘bad’, Briggs?”

Does it matter? It’s BAD, and bad is not good. Are you saying you want this BAD thing to happen?

“How do I know. I don’t even know what it is.”

Are you saying BAD things never happen and that we should do nothing?

“Well, yes, I guess I am saying we should do nothing. I wouldn’t have any idea what to do, since I don’t know what Y is.”

Say, you might be right. Let’s make Y about finance, something about money. BAD things happening to money. You surely don’t want that!

“Maybe I do. Maybe it’s a bad thing happening to your money. That wouldn’t bother me.”

Rude.

“Calm yourself. You must have been watching TV to get yourself in such a state. Just what is this ‘bad’ thing?”

You’re right. I turned it on to see how my predictions on the coronavirus were going and accidentally tuned into The View.

“Can’t say as I feel sorry for you.”

Oh, thanks, Listen: I’m not exactly sure what this BAD THING is. I know it’s unpredictable. It’s something I can’t foresee. Look, unforeseen things have happened in the past that turned into really BAD things. Right?

“Certainly.”

What I want to do is protect against these kinds of things. I’m risk averse. I need to do something!

“But you can’t do anything else you know what it is you’re protecting against. You can’t move to cash in case the bad thing is the government devaluing currency because of some imagined crisis. It’s happened before, you know. You can’t buy stock because there could be a crash. You can’t buy Treasury notes because there could be a war. If anything can happen, then by your own logic any thing can happen. And any thing is a lot of possible things.”

Well this BAD THING has fat tails!

“So? You’re the probability expert. You explain to me what that means. I can’t see it means a damned thing.”

Well…wait a minute…wait a moment…hold up…the effects of TV are fading…fading…gone! I can see things as they really are again! Whew. Thanks.

“You’re welcome.”

I see it all now. You really can’t predict the unpredictable. Some have tried it, though, trying to bootstrap themselves from nothing to something by slathering on nothing a thick layer of mathematical equations.

Look. Here’s what you can do. You can gather events which you have classified as BAD. Stock market crashes, bank runs, currency failures and the like. These take place at certain times. There will be some number of them, and from them you can guess some kind of BAD THING rate, and then you can use that rate, along with some “fat-tailed” distribution for the costs of the BAD, and use it to estimate risk.

But! But, it doesn’t work. It only works if the future resembles the past in all the conditions—the causal conditions. If nothing changes in the causes of stock price changes, then you can use a fat-tailed model to predict extremes. Which every other smart money player is already doing.

If causes change, then your model is useless or will fail you. If you know what the causes are that change, you might be able to adjust your model. But maybe not: depends on the nature of those causes. And if you don’t know what causes change, then you can’t. That’s the idea of “black swans.” Nobody saw them coming, or could. The unknown unknowns (in causes) creep up on you and strike.

“Okay, I get it. Shit happens. We can’t use the knowledge that bad things happen to prepare for bad things about which we know nothing. Don’t be upset, but that seems a bit trivial to me.”

Well, it is, yes.

“So why’s all this black swan stuff so popular, then?”

Beyond the obvious fixes in crude models, replacing normal distribution with ones better in line with Reality, it’s because people think they’ll find a way around all this and find just the right equation that will allow them to see the BAD coming.

“Might as well read the squid ink blots on your plate. That would do just as good.”

You said it.

To support this site and its wholly independent host using credit card or PayPal (in any amount) click here

Categories: Statistics

9 replies »

  1. What’s with all this “fat-tailed”? Are we fat-shaming here? Or is fat-tailed not a bad thing but a good and we are encouraging unhealty tails? 🙂

    If it were not for the unreasoned, emotional, total-loser belief in the ability to predict the unpredictable, we wouldn’t have Schiff saying Trump is going to sell Alaska to the Russians….oh, that was the TV again….. Honestly, humans were not screw-ups before emotions like greed and envy entered but since then, without the ability to terrorize the population with black swans, or aliens or CO2 or RUSSIANS!, or, or or or or how in the world would the evil dictators get to the top? If people understood what you write, there never would be a Hitler, or a Mao, or Marx. It’s fear and the need to predict the unpredictable that created these monsters. You’re messing with history here. Go back to the View and see what crazy things come out of Whoopie’s mouth. You’ll be fat, dumb and happy like much of America in no time.

  2. Death. Death is what’s going to happen. The probability is 100%. Father Time is undefeated. The Big Croak is coming.

    What can you do about it? Nothing. You could make a last will and testament, but that will not forestall the inevitable. You could find Jesus. There might be an afterlife. But maybe not.

    Death. That’s the scary thing. Death is the Black Swan with bony hands and a scythe.

    So carpe diem, live for today, smell the roses while you can. Be a little nicer. Try to find someone or something to love. Love is the answer. You know that, for sure. And find Jesus anyway, whether there’s a payoff or not. You’ll feel better.

  3. Well conceived, well organized, well written. I like it as a dialog, provides continuity.

    An objection that nobody has raised is “But I can predict BAD THINGS with my whiz-bang computer program!” What would your interlocutor say about that?

    w.

  4. Great read!! Thoroughly enjoyed!! Will pass it on!!!

    Thank you, Uncle Mike. “Love is the answer.”

    And God is Love. And Jesus is His only begotten Son. And we can be God’ adopted sons and daughters when we belong to Jesus!!! An absolute surety. We have God’s Word on it!!!

    God bless, C-Marie

    God bless, C-Marie

  5. “So why’s all this black swan stuff so popular, then?”

    Charismatic charlatans with big mouths who have buddies in popular media is why.

    That’s the reason for a lot of ridiculous things.

    Arm yourself with informed skepticism. Then prepare for charlatan attacks!

  6. If this is about Taleb, one his original complaints was that distributions of bad things happening in real life were nog Gaussian, but fat-tailed (which is not a distribution, but anyway). So people would generally under-estimate the change an event ‘in’ the fat tail would occur, if they were doing the estimation the proper, scientific, way.

    Wasn’t he making money by making small bets on bad fat-taily things happening that would pay out a lot if they happened?

Leave a Reply

Your email address will not be published. Required fields are marked *