Prof. Jayanth R. Varma’s Financial Markets Blog

A blog on financial markets and their regulation

Inverting the intermediary theory of asset pricing

In the last few years, the intermediary theory of asset pricing has emerged as a single factor model of asset pricing that does as well as the standard four factor model and thus subsumes the size, value and momentum factors (Adrian, T., Etula, E., & Muir, T. (2014). Financial intermediaries and the cross‐section of asset returns. The Journal of Finance, 69(6), 2557-2596). The theoretical justification for this model is that since financial intermediaries are the marginal buyers of many assets, their marginal value of wealth is a more relevant stochastic discount factor than that of a representative consumer. Though the idea that leverage is a good proxy for marginal value of wealth strains credulity, the empirical results seem quite strong, and there is some case to be made that the shadow price of a leverage constraint is related to the marginal value of wealth.

I see two problems with this. First of all, the major risk factors (like Momentum, Value, Carry and BAB) have been demonstrated in two centuries of data (1799-2016) from across all major world markets (Baltussen, Guido and Swinkels, Laurens and van Vliet, Pim, Global Factor Premiums (January 31, 2019). Available at SSRN: It is evident that the structure of financial intermediation has changed beyond recognition over the last two centuries; for example, 19th Century giants like the Rothschilds operated with far lower levels of leverage than modern security dealers, and were in fact more principals than intermediaries. If the risk factors are solely due to intermediary leverage constraints, I would not expect to see such strong Sharpe ratios for the risk factors in the 19th Century data.

Second, there is a vertical split within the intermediary theory itself. He, Kelly and Manela presented a competing theory (Intermediary asset pricing: New evidence from many asset classes. Journal of Financial Economics, 2017, 126(1), 1-35) with drastically different results. I sometimes joke that Adrian, Etula & Muir (AEM) and He, Kelly & Manela (HKM) refute each other and so there is nothing more to be said. The first direct contradiction is that AEM find a positive price of risk for leverage, while HKM find a positive price of risk for the capital ratio (which is the reciprocal of leverage). Second, HKM get their nice results when they measure capital of the primary dealers at the holding company level unlike AEM who measure security dealer leverage at the unit level. Finally, AEM find book leverage to be more important, but for HKM, it is the market value capital ratio that is relevant.

I am veering around to the view that risk factors are not priced because of intermediary leverage constraints, but it is the other way around. Factor risk premiums have very long and deep drawdowns (for India, the drawdown plots are available at As Cliff Asness put it,

I say “This strategy works.” I mean “in the cowardly statistician fashion.” It works two out of three years for a hundred years. We get small p-values, large t-statistics, if anyone likes those kind of numbers out there. We’re reasonably sure the average return is positive. It has horrible streaks within that of not working. If your car worked like this, you’d fire your mechanic, if it worked like I use that word.

So it is easier to harvest factor premiums if you are gambling with other people’s money especially with a taxpayer backstop for extreme tail events. Since Too Big to Fail (TBTF) banks are ideal candidates for doing this, you could well see significant correlations between the factors and the capital/leverage of these banks, but these correlations might be very sensitive to the measurement procedures that you use. In short, perhaps, we need to invert the intermediary theory of asset pricing.


2 responses to “Inverting the intermediary theory of asset pricing

  1. Josephapape March 22, 2019 at 8:41 pm

    fnmi in fedex

  2. cynicalmeblog March 26, 2019 at 2:24 pm

    Dear Sir, Thanks for the post. I was not aware of this strand of the literature.
    Off late, the whole space of FF factors has come under attack and is being questioned …

    1) Jegadeesh, N., Noh, J., Pukthuanthong, K., Roll, R. and Wang, J., 2019. Empirical tests of asset pricing models with individual assets: Resolving the errors-in-variables bias in risk premium estimation. Journal of Financial Economics.
    2) Raponi, V., Robotti, C. and Zaffaroni, P., 2017. Testing beta-pricing models using large cross-sections.

    I just wonder if too much attention and effort are being put into explaining the anomalies or factors which may not exist in a first place and this may be the reason why finding confounding results are common.

    Another question which has bothered me is if the Original FF paper used Momentum, Value, and Market return because of limitation in computation power and data at that time? Apriori there is no reason to start with these three factors. History repeated with higher computing power/data could have yielded different factors?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: