Hidden Factors Reshape Cryptocurrency Risk Premia

If crypto prices seem to dance to their own beat, this paper says some of that music comes from hidden players. Using weekly data on a broad set of large cryptocurrencies, the author estimates how returns are paid for risk with the Giglio–Xiu three-pass approach, which can allow for omitted latent factors. The headline result is that expected crypto returns load not only on crypto-specific factors, but also on selected equity-industry factors tied to technology and profitability. That points to a tighter link between crypto markets and traditional finance than earlier work suggested. The paper also introduces three non-tradable state variables: Fear & Greed, which tracks investor sentiment; the Altcoin Season Index, which captures speculative rotation; and hacked value scaled by market capitalization, a measure of security shocks. Compared with conventional Fama–MacBeth estimates, the latent-factor approach gives materially different premia for key factors. The takeaway is simple: in crypto asset pricing, ignoring hidden risks can change the story about what is really being rewarded.

Bitcoin does not trade in a bubble. Weekly data from large cryptocurrencies show links to both crypto-only forces and selected stock-market factors. The surprising stock ties point to technology and profitability. Those are equity-industry factors, not coin-specific ones. That means a token can carry more than one kind of risk. It can move with crypto mood. It can also move with parts of the stock market. The pricing test asks what risk gets paid. That matters if you want to know whether a coin is cheap, expensive, or just noisy. It also matters if crypto and traditional finance are growing closer.

When crypto starts borrowing the stock market's mood

The main result is simple, but it is not what older crypto stories implied. The Giglio–Xiu three-pass approach is a three-step pricing test. It lets hidden risks sit beside known ones. Using weekly data on a broad set of large cryptocurrencies, the model finds loadings on both crypto-specific factors and selected equity-industry factors. The strongest stock ties point to technology and profitability. That suggests tighter contact between crypto markets and traditional markets. The study also adds three non-tradable state variables. They are investor mood, speculative rotation, and security shocks. Relative to conventional Fama-MacBeth estimates, the risk prices change in a material way.

How the hidden-factor test works

The three-pass method works like a careful sort. First, it measures how each coin moves with known factors. Then it looks for shared leftovers across coins. Those leftovers stand in for hidden forces. Last, it turns those exposures into risk premia, which are the extra returns investors ask for when they bear a risk. Fama-MacBeth is a common two-step pricing test. It estimates risk prices from cross-sections of returns. This paper keeps that idea, but it lets hidden factors stay in the room. That matters when known factors do not explain the whole pattern.

Why the old crypto story now looks too small

Crypto once looked cut off from stocks and other old markets. Earlier work found little exposure to common stock, currency, or commodity factors for Bitcoin, Ethereum, and Ripple. This paper points to a newer world. Crypto now seems more joined to traditional finance. The non-tradable signals sharpen that view. Fear & Greed captures sentiment. Altcoin Season Index captures rotation into riskier coins. Hacked value captures security loss. None of these are assets you can buy. All of them can still help explain return differences across coins.

Where the surprise leads next

The sharpest consequence is a narrower room for a pure crypto-only price story. A model that skips tech and profit links can miss part of the risk bill. That changes how large coins look when you price them. The next test is the same three-pass setup on later weekly samples from the same large-coin universe. If the stock-style links hold there, the hidden bridge is not a one-off quirk. It is part of how crypto now trades.