class

[class @classpath]

So when the CTA's sell at high prices, who is buying? I dont know, but it does look a bit ponzi-like. An uncoordinated pump and dump.

@choffstein @ScottPh77711570 Would love your thoughts

[class @classpath]

Is trend still your friend? https://t.co/YlsWRQGLz3

The authors talk about the price impact loop of trend-followers. Price up -> TF's buy -> price higher... They say an important part of trend PnL is precisely this price-impact loop.

[class @classpath]

@VivekVRao1 Awesome stuff Vivek.

[class @classpath]

@juke10x @therobotjames Its mostly just beta. Only 12.5bps of alpha/mo.

[class @classpath]

Smoothing usually decreases cov(a,r), as it weights past (less relevant) data more. But, it will also decrease σ_a (noise in your alpha). As long as the σ_a decreases more than cov(a,r), your smoother will boost ic.

TLDR: Smooth alphas to boost ic, let optimizer handle turnover

[class @classpath]

As far as I understand. Smoothing alphas can:

1) Reduce turnover
2) Increase ic

For 1, as ceph points out, an optimizer can do this (and probably should). But an optimizer cannot do 2.

why?

a=alpha, r=return
ic(a, r) = cov(a,r) / (σ_a * σ_r)

[Robot James 🤖🏖 @therobotjames]

@macrocephalopod @NewRiverInvest @robertmartin88 @witchqueendot @stevehouf @Mtrl_Scientist 👋

Say you've got some volatile alpha that explains step ahead returns.

And you wanna smooth it cos real world.

And you observe decay in IC for increasing half life of smoothing. But you also prefer the more autocorrelated alpha.

How do you think about the tradeoff?

[class @classpath]

@SystemicStratHL To push back on the macro view a bit if you look at S&P's forward PE, its ~21x today, which isn't cheap, but not particularly high either.

[class @classpath]

@stalequant By "sees" the TWAP do you mean with public or non-public information?

[class @classpath]

@imotw2 We know thats ur burner otw

[class @classpath]

@midpricedog What tool?

[class @classpath]

@m1ha313 @systematicls cache consistency like making sure its not stale?

[class @classpath]

@midpricedog Why is this simpsons paradox? Are men applying to easier departments to get into?

[class @classpath]

@sthlmlondon sure

[class @classpath]

I doubted this claim so I priced the second contract (btc hits 100k in Nov) with a monte carlo.

Polymarket: 40% for YES
My model: 40.7% for YES

[class @classpath]

@midpricedog his equity curve so good he puts it on a log scale. Link: https://t.co/YiDSheWgjQ

[class @classpath]

@GryptoGoon how come?

[class @classpath]

@imotw2 Its always a good day when otw posts trading content

[class @classpath]

@xmgnr chin up king

[class @classpath]

@DegenQuant @yenwod_ banger

[class @classpath]

@macrocephalopod Youll be missed

[class @classpath]

@stalequant How does credora calculate p(default)? If its a function of APY then that would explain the good r^2 but not necessarily mean its correctly priced.