Dear Investing Community inside 2p2,

this has probably been asked before, but I cannot find sufficient stuff on the internet, so I turn to you, my dearest, oldest poker community:

Background:
So, I was an online poker professional ten years ago, then turned software engineer, then turned quant, now getting tired of that.
In my time, "winning" meant: having a significant (>4bb/100) winrate over at least 100k hands.

Problem
So, I thought to myself: Why not try trading a bit? Portfolio Management is trivial, so why not try exactly what I did before: dealing with other people's expectations`?
And become a trader?

So I read a book by Nassim Taleb about Trading: "Dynamic Hedging". And understood nothing.
Except: He wrote: "I am a trader with 70,000 trades on my belt life-time." And he meant: "So I know what I'm talking about! I am mad experienced!"

As a former poker pro this seems kind of ridiculous to me: You play 100k hands a month, with a 25% VPIP, 50/50/50 (to make it easy) WtT, WtR, WtSD, you have a
whopping 100k + 25k + 12.5k + 6.75k + Reraise/Fold decisions to make within a month. Maybe 200k?

Question
A man who is so totally fond of himself understanding variance - Taleb - how can he come to the conclusion that his 70k decisions lifetime constitute to much?
How can I even remotely relate that to my understanding of variance as a former poker pro?

First off Taleb is an egotisical dickhead. That said Fooled By Randomness is one of the most important books ever written Anything else he has written is probably not worth reading. Poker does not equal trading.

You can sit at a poker table forever and thus play many hands. You can stare at the market and wait for a trading opportunity for a while. I was a market maker so I made many more than 70K trades in my lifetime because it was my job. Now that I am not "providing liquidity" I don't trade nearly as often. More trades (usually out of boredom) isn't a good thing generally. I never read the book you mention but since Taleb is such an insufferable prick I can't help but think it is off putting.

Hahaha!

Thank you very much! That's exactly the lore I love to hear in this forum.

Why don't you think, "providing liquidity", "making markets", just: "making more trades" makes the variance smaller?
Where should I adjust my thinking?

My - probably naive - thinking: Betting on whatever is a noob game. In the long run only the guy who has the edge is
the guy trading high frequency.
Everything else is just variance.

That's just me from a former poker pro perspective.

Why is that so wrong?

I am probably very naive, but that's why I am asking these questions right here.

it's very different. When you play poker you want to play a lot cause you want to put volume on it. When trading you don't want to open too many trades because in lower timeframes is all about noise, add the spread plus fees and you have zero chance to beat lower timeframes. Lots of traders say the money is on daily-weekly charts.

Also is a very shady world, too many people say they win money but it's almost impossible to probe, also lots of traders win money in a bull market but but ironically they would have won more just with the buy&hold strategy cause in a bull markets everyone makes money.

Variance in trading and variance in poker are massively different .

Luck variance in poker is equal for everyone once u play tremendous amount of hands .

In trading you make a bad trade , u can’t count on luck to bail you out .
( unless u consider others making bigger mistakes then you being luck )
So there is no points to try mitigate variance by increasing volume trading .

Warren buffets is not considered a day trader and he probably is the best track record investor ever or close to .

In investing you have the option to only play aces because if you don’t play , no blinds comes in to eat your saving (disregarding inflation to make a point ).

In poker you have blinds that oblige you to play more hands to beat the blinds cost .

If no blind existed in poker , it would be a mistake to play any hands other than AA .

A trade is more analogous to a poker session than a single hand.

The simplest way to think about this is that even the very best poker player's edge on a per-hand basis is extrememly small. Hence, you need an extremely large number of hands to overcome variance. Virtually no human trader/investor operates with such low return expectations on a per-trade basis.

Let's consider a hypothetical +5BB/100 player, who plays with an average stack size of 200BB. This player earns 0.05BB per hand, despite risking 200BB every hand. That's the equivalent of investing $40,000 with the expectation of earning $1 per trade. Now, typically, even when you lose, you don't lose your entire stack, so the effective amount at risk per hand is much smaller, but this is also true of trading, you typically don't lose everything when the trade doesn't work out for you.

For most trades and investments, time is also a significant component - if you buy something and sell it 100 days later, you effectively made tons of independent bets to keep the investment - the decision to hold the investment from 1pm to 2pm is largely independent from the decision to hold the investment from 2pm to 3pm, from a pure risk-reward perspective.

Time is your measurement, not trades made like hands played. When I was trading more actively than I am now I would break it down to days, weeks and months. I never looked at trade by trade types of results. Obviously, days have more variance than weeks and weeks have more variance than months.

Trade by trade I had more losers than winners but with my approach my winners were bigger than my losers. Day by day winning days were still slightly below losing days. But week by week the tide turned and I had a much higher percentage of winning weeks. Monthly a down month was pretty much non-existent.

Now I am just playing around with option swing trades and I do a lot of far out calendar spreads and condors and butterflies. They can often take a few months to materialize.

If you look at the formula for variance and think about it for a bit you can see what really matters is that the magnitude of the outcomes are similar (ie: because of the squaring).

So for poker you need such large sample sizes to get a representative sample of the few % of hands that really effect the variance; all those 1000's of hands where you fold preflop and lose a tiny amount make almost no contribution to the final variance per hand.

This is an interesting old thread that shows this:

https://forumserver.twoplustwo.com/2...d-not-1410413/

Iirc, there were 1/2 a billion samples but just a handful of huge bets by "whales" completely swamped the variance and bootstrapping showed the chance of the "bad luck" from the investors perspective was about a 1/20 event.

Juk

Are those option trading lingo? Never heard of those term before.

You should check out a site called Google. You can search stuff like "condor options" and "butterfly options" and it will show you results from other pages on the internet.

grunching here...

i've seen strategies from the credit crisis that would have killed it.

BUT....

some of those strategies would have consistently lost money (on some rolling period, say 3 years) since the early 1960's. that's the longest i had data back for them.

so you would have had angry clients every year for 46 years............

it would be interesting if both players (perhaps both voting on paper simultaneously) could agree to take the EV of their respective hands before an all-in battle... reduce 80--20% variance for example.

pains me to say this, but sklansky had a good idea...... he said (paraphrase, and i made up the exact numbers), "sports betting sites/casinos should have you bet point spread on 8 games......... but the max payout and max loss is based on a range of 2 to 6 winning bets... just eliminate 1, 2, 7, 8 winning bets (floor and collar, in wall street option lingo).

i am going to model the variance of this idea.

as a follow-up,

i am having great success on many free sites (yahoo, bigcharts, barcharts and more) getting two stocks on the same chart i.e. each of their price charts.

BUT,

still shutout on getting a line of "relative strength" for the 2 price series.

i am dinosaur but i remember this being much easier years ago when there were fewer choices.