Kelly Criterion

Amateur investor. (Mostly I reply rather than post.)
#StandWithUkraine

[Kelly Criterion @JohnsCriterion]

@orrdavid daily vol is higher (than random walk would predict) and the difference likely increases with lower liquidity. So a Sharpe ratio based on daily prices might be biased against a portfolio of (long or short) smaller cap stocks.

[Kelly Criterion @JohnsCriterion]

@orrdavid Semiconductor design is also software-like.

Semis (imo) can be viewed from the same lens as software:
moats originally held by (1) man-hours of brain labor and by (2) platform network (coordination) effects are now to be held almost entirely by (2).

[Kelly Criterion @JohnsCriterion]

@CliffordAsness I believe it is just possible. break even is when the percent basis B and marginal tax rate T are at T = 1/(2-B). Currently no one in U.S. has T>.5 for L.T. gains. But for S.T. gains in NY/CA it can happen if forced out of a position (i.e. if $ABVX had been bought out for $).

[Kelly Criterion @JohnsCriterion]

@CarioCapital @BillAckman one way to think of it in round numbers: if the expected gain from copying him is 10%, then the 2% fee is paid for perpetually by earnings of 10% on the 20% nav discount. If the expected gain is 8% then 25% discount is needed.
(ignoring tracking loss and .2% expenses)

[Kelly Criterion @JohnsCriterion]

@TheStalwart nvda

[Kelly Criterion @JohnsCriterion]

@orrdavid Yes, however I find timing your entry can reduce your transaction costs (defined very broadly).

Most of these trade-timing ideas are not macro, but some are (like opposing the market’s reaction to the administration’s words as opposed to their actions).

[Kelly Criterion @JohnsCriterion]

@orrdavid The Kelly violating problem when leverage is >1 is missing some outlier factor causing surprise correlation. Likely the #1 factor is a market liquidity event with B>1. #2 is an auto-liquidity event. Just saying: some correlations are serious red flags and others are not.

[Kelly Criterion @JohnsCriterion]

@GaryHaubold @ArmandDAngour by “reduce” I mean cross them off the list — don’t kill them.

[Kelly Criterion @JohnsCriterion]

@GaryHaubold @ArmandDAngour Perhaps the baseline was that the whole group of 100 or so soldiers were guilty. Reduce the guilty group by 90% by sampling (decimate it), now 10 guilty.

[Kelly Criterion @JohnsCriterion]

compare trades with Returns and Probabilities
R1:{30%,10%,-10%,-20%,-70%} P1{.3,.3,.3,.08,.02}
R2:{29%,9%,-11%,-21%,-71%} P2{.3,.3,.3,.1,0}
(R2 has tail risk insurance)
Kelly Fraction F maximizes CAGR if bet repeatedly
F(R1,P1)= .97 with G=3.44%
F(R2,P2)= 1.98 with G=5.96%

[Kelly Criterion @JohnsCriterion]

Why buy Puts or sell short? If you have an edge, insurance against catastrophic loss can improve the optimal gain (by increasing the rational leverage).
This isn’t because insurance is cheap (the prob of catastrophic loss is estimated by the price of the insurance).
example:

[Kelly Criterion @JohnsCriterion]

@dampedspring for example if you hold 50 years before selling, you net 108x which is a cagr of 9.8% or equivalent to a 2% yearly tax on gains.

[Kelly Criterion @JohnsCriterion]

@dampedspring over the long term, if you go decades without selling, tax approaches 0%/yr (final tax is like losing last 2.3 yrs).

Suck case is 10% realized as s.t. gains each year nets only 5.9%/yr (or as l.t. nets 7.7%/yr).

This is not including state taxes (9.9% l.t. in WA).

[Kelly Criterion @JohnsCriterion]

@orrdavid that was a long way to say I try to let my winners ride.

[Kelly Criterion @JohnsCriterion]

@orrdavid for my style, in tech at least, you don’t sell your winners due only to price getting a year or so ahead of growth. Losers can be on a hair trigger— I just tell myself to buy it later when it is on its way up.

[Kelly Criterion @JohnsCriterion]

@orrdavid Alternative view:

Adverse selection hurts at both entry and exit (it is worse for an amateur like myself with less edge and fewer ideas). This causes hysteresis in conviction between buy and sale trades

[Kelly Criterion @JohnsCriterion]

@Biohazard3737 please also pay in stock structured to avoid immediate capital gain.

[Kelly Criterion @JohnsCriterion]

@tradingsssss idm is an understatement!

For the U.S. owners, a stock transaction “merger” could avoid immediate realization of the capital gain and also potentially avoid our much higher short term rate. (but I’m no expert and there are other requirements)

[Kelly Criterion @JohnsCriterion]

@siyul depends what you mean by often. Often it is luck combined with adequate position sizing.

However it would be bold to claim (if you are) that the Bayes Update of skill given risk adjusted performance is backwards this year.

[Kelly Criterion @JohnsCriterion]

@orrdavid May not work if you have (a team which includes) an activist short seller.