This week I discovered the vibe quant movement (or rather, it discovered me). People using LLMs to find trading strategies, validate them, and put them into production. The pitch is seductive: the LLM reads the literature, implements the ideas, backtests them, and you just supervise. I think this approach is going to cost people a lot of money. And worse, its going to prevent them from

The goal of our article is to examine the long-term relationship between small value and large growth stocks using more than 100 years of data and test whether the spread between small value and large growth portfolios shows trends that could help investors switch between the two styles. Using the Fama and French 23 and 55 size and book-to-market portfolios, we construct the small value minus

(but that wont always be the case) Over the last 15+ years, diversification (as opposed to market timing) has been a huge drag on Tactical Asset Allocation (TAA) performance, to the tune of 2.1% per year compared to the 60/40 benchmark. That diversification drag has been mostly due to US stock market dominance over almost all other asset classes over that period. TAA must use market

At eight words this is almost certainly* my most viewed and liked tweet ever (although I have nearly 25k followers, so thats 18,000 or so that didn't like it) . Short, pithy, funny; I should retire from my Xmaxxing game right now (just kidding; there are still plenty of gamblers, crypto nuts and MAGA idiots waiting patiently to be educated; and I have a million more dad jokes to inflict on

Why Breaking In Feels So Hard (and Why It Is Still Achievable) If you are trying to get your first role in quantitative finance, you have probably noticed two things: Job descriptions often look intimidating Everyone online seems to have a PhD, a perfect CV, or both That can make the whole path feel inaccessible. In practice, most successful candidates do not "have everything." They are

This study examines the effectiveness of several anomaly-based trading strategies applied to the real estate sector represented by the RlEst index from the FamaFrench 48 industry portfolios. Using monthly data from July 1, 1926, to December 1, 2025, we analyze whether selected strategies are capable of generating superior risk-adjusted returns compared to both the standalone RlEst index and the

My last two articles on AI and trading research got more engagement than almost anything Ive written. More of the Disease, Faster argued that LLMs cant answer the critical question: who pays you and why? AI Will Create Millions of Quants went deeper on the why: AI makes beautiful backtests trivially easy to produce, which means more false discoveries, more overfitting dressed up

Whenever John Ehlers writes about a new indicator, I crack it open and wire it straight into C for the Zorro platform. Or rather, I let ChatGPT do most of the work. The One Euro Filter is a minimalistic, yet surprisingly effective low-latency smoother that reacts instantly to volatility with less lag of the usual adaptive averages. This is achieved by dynamically adapting its time period. This is

In the previous article, we established that the Sharpe ratio is the single most important number in portfolio construction. Variance drag scales with the square of volatility, which means a high-Sharpe portfolio can tolerate leverage, survive decumulation, and compound wealth far more efficiently than a low-Sharpe one. We also hinted at something more subtle: strategies with positively skewed

Quantitative researchers have always sought new methods to extract meaningful signals from noisy financial data. Over the past decade, the field has progressed from linear factor models through gradient-boosting ensembles to recurrent architectures such as LSTMs and GRUs. This article explores the next step in that evolution: the Transformerand asks whether it deserves a place in the

Before moving on to the next series, theres an important point that any algorithmic trader needs to know: the point about APIs. A quant can state the issue in simple terms. Let st denote latent market state and let xt denote observed state at decision time. The strategy trades on xt rather than st. The transformation from st to xt includes vendor capture, transport, buffering,

Quantitative trading is about many things; buying low and selling high, chasing momentum, capturing yield, arbitraging price imbalances the whole lot. But time and time again, it comes back to one idea: The cross-section, the cross-section, the cross-section. Now, if you dont spend the majority of your life in code terminals and academic papers, the word cross-sectional might not mean much.

Applications of machine learning in finance continue to evolve rapidly. In previous posts, we discussed both the uses and the challenges of applying machine learning in financial markets. In this installment, we continue that discussion by highlighting new research on machine learning approaches for pricing complex derivatives and identifying signals that may precede major market downturns.

Traditionally, balanced portfolios rely on the equity and bond risk premia to generate returns. The canonical 60% equity/40% bond allocation has produced strong long-run results, but its short- and medium-horizon experience includes large drawdowns. Those drawdowns matter not only behaviorally but mechanically, through volatility drag: large losses require disproportionately large gains to

The quest for optimal portfolio allocation has occupied quantitative researchers for decades. Markowitz gave us mean-variance optimization in 1952, and since then weve seen Black-Litterman, risk parity, hierarchical risk parity, and countless variations. Yet the fundamental challenge remains: markets are dynamic, regimes shift, and static optimization methods struggle to adapt. What if we

AI makes it easier than ever to build trading strategies. Prompt a model, run a backtest, optimise some parameters, and suddenly youve got a beautiful equity curve staring back at you. It feels like progress. It feels like research. I wrote recently about how AI coding assistants tend to prescribe more of the disease, faster, skipping the learning that makes trading research actually

Regression-based machine learning is suitable for optimizing macro trading signals, particularly for combining multiple trading factors within a strategy. However, due to the low frequency of macroeconomic events and trends, the bias-variance trade-off in machine learning is very steep, meaning model flexibility comes at a high cost of instability. To improve the trade-off, regression-based

How to take a Python financial model from running 150 scenarios in a Lambda function to processing over a million using AWS Step Functions, Batch, and Fargate without managing a single server. From Laptop to a Million Scenarios You've built a financial model in Python. It runs beautifully on your laptop perhaps a portfolio stress-testing model that grinds through a few hundred

For many investors, short-term interest rates are often treated as something the market discovers. In reality, the Federal Reserve has enormous control over how the front end of the yield curve evolves. While textbooks often portray the Feds policy rate as a flexible tool that reacts quickly to economic data, the actual behavior of the Federal Open Market Committee (FOMC) looks very

In the previous post of this series on measuring financial risk, I described the absorption ratio, a measure of financial market fragility based on principal components analysis, introduced in Kritzman et al.1. In this new blog post, I will describe another measure of financial distress called the market rank indicator (MRI), this time related to the notion of condition number2 of a matrix,

This article is a continuation of a series of articles on generating synthetic equities datasets for the purposes of machine learning (ML) model training or synthetic backtesting of systematic trading strategies. We have previously considered the generation of synthetic correlation matrices and the generation of synthetic asset returns via various time series models. In this article we are going

Every day, financial news outlets publish thousands of articles about publicly traded companies. For investors, the obvious question is: does any of it actually matter? If a headline says a company just signed a major contract or passed a clinical trial, should you expect the stock to move the next day? Thanks for reading! Subscribe for free to receive new posts and support my work. This article

In the last article, we built up a conceptual understanding of universe selection: how to find pairs that diverge and converge in a tradeable way. We talked about measuring the thing you actually care about directly, rather than reaching for statistical tests like cointegration that sound perfect but turn out to be unstable in practice. The natural next step is to start trading them the

Even a single basic macroeconomic factor applied to one derivatives market can generate material and consistent long-term risk-adjusted returns. This is illustrated using point-in-time GDP nowcasts in global FX forward markets. The deployment is based on a simple premise: relatively strong economic growth positively affects local-currency FX returns, owing to its support for higher real interest

Selecting mutual funds is one of the most important jobs investors face. Yet the tool everyone reaches for, the Sharpe ratio, quietly assumes something most real people do not have: the ability, and willingness, to borrow at the risk free rate to lever the best fund up or down to their preferred risk level. Once borrowing is realistically restricted, the Sharpe ratio can stop lining up with

No, it's not one of those 'make $$$ easy by trading' courses, it's a dull and tedious one about robust fitting and backtesting. This is the first* time I've taught outside of a university. * and possibly last, we'll see. This could be a one-off opportunity. In person and remote:

n my recent piece on Kronos, I explored how foundation models trained on K-line data are reshaping time series forecasting in finance. That discussion naturally raises a follow-up question that several readers have asked: what about the architecture itself? The Transformer has dominated deep learning for sequence modeling over the past seven years, but a new class of models State-Space Models

This research examines the critical quantitative investment problem of systematic tactical allocation to international equity mandatesspecifically Emerging Markets (EM) and Europe, Australasia, and the Far East (EAFE)amidst conjectured macroeconomic regime transitions. The investigation is precipitated by observable deteriorations in USD hegemony, elevated geopolitical risk premiums, and

In the recent previous article on Correlation Matrix Generation using Object Oriented Python we created a Python object-oriented class hierarchy to develop an extensible, modular tool for generating synthetic correlation matrices. Such matrices can be used to generated synthetic correlated time series models, which can form the basis of realistic synthetic financial datasets. In this article we

Casino gambling with a system where you have the edge is a wonderful teacher for elementary money management. Ed Thorp Ed Thorp is the money manager I admire most. Many people have never heard of him. They should have. In 1961, with Claude Shannon the father of information theory he built the first wearable computer to beat roulette. He wrote Beat the Dealer and proved blackjack

Alternative ETFs sit at an unusual intersection of public-market accessibility and hedge-fund-style investment techniques. They package managed futures, merger arbitrage, and option-based income strategies into exchange-traded products, yet they remain thinly traded and relatively niche compared to mainstream equity or bond ETFs. This combination makes them intriguing: they offer exposure to

In fund research, the input that matters most is simple: what stocks are inside the funds right now, and in what weight. Without that look-through layer, fund momentum, category rotation, or risk exposure becomes a label-driven proxy. We are going to construct and execute a systematic fund scraping operation. The explicit objective of this architecture is to expose what underlying stocks these

Momentum has been studied extensively across equities, commodities, and other asset classes, with well-documented evidence of cross-sectional and time-series continuation effects. More recently, an emerging line of research has shifted attention to momentum in option returns, examining whether derivative markets exhibit their own systematic return patterns. In this post, we review the latest

The quant finance industry has spent decades building specialized models for every conceivable forecasting task: GARCH variants for volatility, ARIMA for mean reversion, Kalman filters for state estimation, and countless proprietary approaches for statistical arbitrage. Weve become remarkably good at squeezing insights from limited data, optimizing hyperparameters on in-sample windows, and

As we all know, backtesting is not a research tool, but the very end of your research pipeline. If you want to evaluate if a given set of signals is predictive for returns, you can do this more clearly and directly by regressing returns on the signals or measuring their correlations. But how strong do those correlations need to be for the signals to be good enough? A popular heuristic

When Is a Signal Good Enough? As we all know, backtesting is not a research tool, but the very end of your research pipeline. If you want to evaluate if a given signal is predictive for returns , you can do this more clearly and directly by regressing on or measuring their correlation. But how strong does that correlation need to be for the signal to be good enough? A popular heuristic

CAPE Ratios and Long-Term Returns Rui Ma (La Trobe University), et al. January 2026 We demonstrate that 10-year equity market returns are considerably more predictable in relation to price-earnings ratios than previously thought. The traditional approach involves relating the current index price level, based on current index components, to the index earnings of previous years, calculated using