cheesecake (@Cheesecake_Lord) profile picture
purgatory
Joined August 2011
Photo by cheesecake (@Cheesecake_Lord)
We're giving away a Hogwarts Legacy custom wrapped NVIDIA GeForce RTX 5090 to celebrate its official release with DLSS 5 🪄 To enter: 🟢 Share this post 🟢 Comment #RTXON
12,576
11,228
593
11,200
857,733
We're giving away a Hogwarts Legacy custom wrapped NVIDIA GeForce RTX 5090 to celebrate its official release with DLSS 5 🪄 To enter: 🟢 Share this post 🟢 Comment #RTXON Terms and Conditions: nvidia.com/en-us/geforce/con…
84,373
71,862
7,696
66,836
6,068,975
cheesecake retweeted
JUST IN: Jim Cramer says investors should sell data center stocks "no matter what"
Readers added context they thought people might want to know
Jim Cramer did not say to sell data center stocks "no matter what"; he cautioned against overexposure to tech broadly while stating "I like data centers so much that I wish I lived in one." youtube.com/watch?v=Ztauhn… finance.yahoo.com/markets/stocks… youtube.com/watch?v=kU3HZz…
256
123
163
1,904
592,628
cheesecake retweeted
JUST IN: London Stock Exchange to launch 24-hour stock trading
127
246
66
2,744
217,450
cheesecake retweeted
JUST IN: Jim Cramer declares the market is currently “miserable.”
252
90
125
1,465
587,310
cheesecake retweeted
giveaway time! win a Claude Max 5x subscription to enter > follow @BleapApp > rt and like this post winner will be selected in 24 hours (btw, you get 20% cashback when using a Bleap card for your claude subscription anyway)
211
633
21
1,009
77,093
cheesecake retweeted
🤍 Valentine’s Day doesn’t need much. Just the right power and the right moment. We’re giving away an Osmo Pocket 3! How to enter: 1. Follow @djipowerglobal 2. Like, comment, and repost this post Campaign: Feb 13–Mar 2, 2026 Winner announced: Mar 2, 2026 Good luck!
1,334
1,094
42
1,424
4,299,535
cheesecake retweeted
This is another crypto signal that works! I am, of course, talking about the mighty value signal, but in crypto! It is mightily easy to construct, and is essentially proxied by a single estimator. As usual, happy to share the article with some comments + reweets on this post!
14
15
62
13,910
cheesecake retweeted
The Problems With PCA We've all had pipedreams of extracting predictive features from PCA, but they've mostly fell flat. Why? I think in part it's because the factors have no stable meaning. PCA extracts whatever linear combination explains the most variance. Secondly, the stocks themselves have no stable identity either. With this, it rewards memorizing idiosyncratic patterns specific to each stock. You have enough flexibility to fit noise rather than signal. Today's article is about a different form of PCA that overcomes this and actually is able to extract predictive features! You guys know the shtick, comment and retweet for a draw of the free article.
19
23
59
16,811
cheesecake retweeted
There are signals that just "makes sense", and when you find them, you need to do almost no kind of backtesting and fitting. This is one such example. Once you understand this signal, you will go, "That totally makes sense, why didn't I think of that?!" Most people don't understand what it means to be "creative". Every signal must be momentum or mean reversion or *shudders* pairs trading. I hope to share with everyone that good signals can be found by thinking deeply about behavioral effects and existing algorithmic signals. Random comments and retweets will get this article for free.
58
41
2
159
22,972
cheesecake retweeted
You read about index rebalance strategies everywhere, and it seems like a hedge fund darling, why? It's one of the few systematic* strategies that are extremely scalable (you can run BILLIONS in capacity), and where the economic rationale for working is easily understood. Index rebalance is all about predicting how much dollar flow is going into a stock relative to its average daily traded dollar volume. A high dollar flow relative to its usual traded dollar volume will mean the stock price will go up, and vice versa. It is beautifully simple conceptually, but VERY difficult to get right practically. A lot of PMs talk a big game about index rebal, but actually getting it right and managing your risks is NOT easy. Especially when so much of your gross returns can show up as momentum and short interest factors! --- I write below about a foundational methodology, a starting point for thinking systematically about index changes. You’ll build more sophisticated signals over time, but this framework captures the core mechanics. This methodology on systematic index change prediction crystallizes an approach many quant shops have traded for years. Comment AND retweet on THIS post to get a chance for a free article!
22
22
116
21,161
cheesecake retweeted
Let’s say you have many signals, and they all have different turnover profiles, and you are interested in combining them into an optimal portfolio. How should you begin thinking about this? One simple heuristic is to weigh your signals by their IR. The higher the IR, the higher their weights. That’s a good start, but there’s a few problems… IR is an implicit bet that your signals have persistent performance AND IR doesn’t account for realizability/tradeability, which is a function of your turnover. One will realize that the higher the turnover of your signal, the harder it is to monetize it and therefore the less weights it should actually get. So, what is a good heuristic for this? I talk a little more about this in the following post:
2
2
46
7,724
cheesecake retweeted
Almost all of us will deal with tick data at some point in our careers. Even a MFT book will likely deal with tick data to transform it into some aggregated form. Unlike aggregated price data, it can be tricky to deal with the large volume of data in real time. Most practitioners will implement some kind of filter for both historical/live to catch any outliers. There are "goldilocks" methods that allow us to approximate "optimal filtering" when dealing with millions and millions of rows, and this article discusses exactly one of them! As usual, happy to share the article with some who retweet and comment on THIS post!
25
25
1
106
22,826
cheesecake retweeted
If you join a statarb team, you are essentially trying to run variants of mean-reversion (MR) strategies. The catch is this that traditional MR (betting that gross returns will revert) is negative after-cost since 2010. So, how have statarb teams adapted? Simple, they've learnt how to make mean-reversion strategies in a different "space". "Gross returns" are an example of a statarb "space". Today, i want to talk about a really novel "space" that still produces meaningful performance even till today. Perhaps some of you may have a clue, but the hint is that it involves rethinking factor decomposition! If you'd like to find out more, head over to the bad place AND as usual, I will give out random free reads to retweets + comments on THIS POST.
32
42
1
228
85,522
cheesecake retweeted
You've heard of crypto trend signals. Have you wondered how they were created? Today, I actually show you guys how to create a performant, scalable trend signal. Hint: It's going to be a smart ensemble of weak trend estimators. Randomly giving out free articles for retweets!
10
54
2
104
16,857
cheesecake retweeted
You will often hear practitioners say that when you generate forecasts for instruments, what REALLY matters is the RANK of your forecasts, not the actual MAGNITUDE. Why? It's because in a cross-sectional strategy where you are going long some instruments and short some instruments, you make money from getting the RANKING right, even if you can get the magnitude very wrong (really, I show an example later)! So most systematic long short flows look like this: 1. You generate some forecasts using some ml model 2. You rank the forecasts 3. You go long/short the top/bottom decile Here's today's problem... your ml model is trained on MSE, which is a pointwise estimate that doesn't give 2 shits about ranking! So you can have fairly accurate forecasts and still get ranking wrong (and lose money)! --- How do we solve this? Read on at the bad place to find out more! As usual, happy to give out free reads to anyone who retweets! Have a great day!
17
34
2
110
18,726
cheesecake retweeted
You have produced multiple forecasts over multiple horizons, how do you trade them optimally? These are the things your PM (the monetization team) is trying to solve! When you have forecasts at multiple horizons, the natural instinct is to combine them somehow and trade toward the combined target. Perhaps you'll try equal weights. Maybe returns-weighted. Maybe volatility-adjusted. All of these are wrong! Want to understand how your PM monetizes multiple forecasts? I'm giving away free articles to random retweets as per usual!
8
50
1
89
17,538
cheesecake retweeted
The context in which people hear about factor models is always about using factor models to hedge. This is ONE side of the story, where you are trying to remove systematic risks from your portfolio. The real pros understand that factor models are a denoising technique. They partition returns into “stuff that does not matter” (systematic factor exposure) and “stuff that does matter” (idiosyncratic returns where alpha lives). The better you model the systematic part, the cleaner the residual and the cleaner the residual, the easier it is to find tradeable alpha. In today's article, I talk about this phenomenon and how to roll your own factor models no matter what markets you are trading! As usual, I will be giving out the articles for free to some retweets!
11
40
1
111
18,652