Matt Brigida

How much was the Google Quantum AI mention worth to ALGO? Roughly $302M, after backing out the broader crypto market move. Algorand already had the technology. The paper got the market to price it.

Update: ALGO's alpha hit 12% today alone. Total announcement-specific return now 39%, net of the crypto market. 4 days post-Google Quantum AI paper and still pricing it in. Textbook post-announcement drift. Other positive news increasingly a factor. …s-announcement-event-study.vercel.app

ALGO rallied 26% around the Google Quantum AI whitepaper (after stripping out the crypto market factor effect). Here is an interactive event study where you can choose your own estimation and event windows: …s-announcement-event-study.vercel.app

@artemis Thank you, happy to chat and compare notes anytime.

@mattbrigida2 Matt, would love to chat more and compare notes. Hope you're doing well.

Crypto markets just had one of their worst months in the dataset. The broad market factor fell -23.5% in February. But Artemis’ factor model shows something interesting: • Momentum: +11.3% • Value: +10.9% • Growth: +2.6% In other words, even in a major drawdown, systematic factors still generated positive returns. Full analysis below 👇

@artemis Also since the same names (HNT, Morpho, ARB, OP) show up across factors, I suspect your reported factors are spanned by the market and SMB.

@GovernorHat Exactly. The results suggest attention helps translate fundamentals into performance, so coordinated amplification makes sense.

@mattbrigida2 At the very least, other AF accounts/people should be coordinating to amplify things. This is the lowest of low hanging fruit.

TVL alone doesn’t predict higher crypto returns, but TVL + fast X follower growth might. High minus Low port returns are positive and significant after controls. Evidence for social factors in pricing. matt-brigida.github.io/TVL_X_multisor…

@LoafPickle Your intuition on social factors was correct! I’ll see what I have for data on amm volume.

@mattbrigida2 Can you check TVL + on chain amm volume next?

@GovernorHat Fair point. The goal isnt views today, but better policy tomorrow. Maybe I need that blue check after all?

@mattbrigida2 if social engagement is important, then this isn't looking good.

More data and better measurement may sharpen the effect. Could be a useful factor.

New post: matt-brigida.github.io/TVL_DEX_blog_p… Can TVL and DEX volume activity deliver alpha in crypto? I test this with weekly 2x3 sorts. The DEX activity signal looks encouraging in simple averages, but robust SEs temper the conclusion.

@LoafPickle Absolutely, good idea. I’ll just have to find some decent data on SOV.

@mattbrigida2 Can you try to find a correlation between price action and share of voice?

Some community feedback on my 2025 paper(sciencedirect.com/science/articl…) was that TVL may be relevant in conjunction with other variables. So I began testing 2x3 multisort portfolios. Here are the first results---on TVL and Active Developer multisorts: matt-brigida.github.io/TVL_devs_blog_…

@LoafPickle The method used is to create crypto portfolios based on these variables (the multisort portfolios) and test whether they generate alpha.

@LoafPickle We are testing whether increasing the number of active developers (controlling for TVL) increases crypto prices. We find it doesn’t, but we have promising results with other variables which I’ll post soon.

I also test new non-tradable state variables (e.g., Fear & Greed, Altcoin Season, and hacked value scaled by market cap).

New working paper on arXiv: “Crypto Pricing with Hidden Factors” arxiv.org/abs/2601.07664 The paper estimates risk premia in the cross-section of cryptocurrency returns while allowing for unobserved (latent) risks alongside standard crypto and equity market factors.

Key takeaways: - Accounting for latent risks can materially change estimated premia versus conventional approaches. - Results are consistent with stronger links between crypto and traditional markets.

@artemis Great analysis. One thing that jumps out is the use of equal rather than value-weighted portfolios. Did you also try value-weighted versions, and were the results similar? Equal-weighted long–short portfolios can be very costly to implement, especially in illiquid names.

Equities have lived on factor models for 30+ years. So we tested them in crypto: • Market Beta portfolio: 42% annualized performance • Size (SMB): 53% annualized • Momentum: 75% annualized On diversified portfolios, factors explained ~75% of returns. Full analysis from @covered_call 👇