Recent Posts

Sunday, November 23, 2014

Musings On Top-Down Macro

I wrote a review of the book Investing from the Top Down: A Macro Approach to Capital Markets, by Anthoni Crescenzi (now at Pimco), on the Personal Finance section of my web site. The book is relatively old (published in 2009), but it is one of the few books (that I know of) that cover top-down macro investing. (That is, making asset allocations based on a view on the macroeconomy and aggregate market pricing, as opposed to bottom-up security picking, where you scour corporations' financial statements to find good buys.) The book's main audience is retail investors, but it may be of use for people working in finance that are unfamiliar with top-down macro.

What I find interesting is how under-served this approach to finance is when compared to other methodologies.


In order to make my point more precise, I would argue that:

  • there is little academic research that is of any use, and so analytic tools are relatively primitive;
  • other than macro hedge funds, there are few means to find asset managers that invest in this way, other than those that are constrained within an asset class (for example, emerging markets, currencies, actively managed investment grade bond funds);
  • correspondingly, very few people in finance are actually engaged in thinking about relative performance across asset classes. 

I will give an example based on my experience. It is possible to find analysts who could use sophisticated computer algorithms, huge data sets and some serious mathematics to tell you where the 5-year swap rate should be, conditional on having the levels of the 2-year and 10-year swap rates. (In fact, the analysts may even get results that are within a few basis points of each other.) But if the question is changed to: "what is the fair value of the 5-year swap rate?", analysts have to revert to the equivalent of examining goat entrails.

This is repeated throughout finance. We have answers to esoteric questions, such as how to hedge an exotic derivative. But finance cannot answer useful questions, such as whether shale "fracking" is a bubble or not. Since the usual justification for the social utility of financial theory is that it can be used to best allocate capital, this is troubling.

It is easy to understand the source of the problem: predicting the future is hard. Moreover, mainstream macroeconomic theory has retreated into a cocoon which is cut off from reality.  For example, you need teams of Ph.D. economists, armed with stochastic calculus, to develop models that tell us that recessions are caused by people purchasing less stuff. Given the lack of guidance from academia, finance practitioners are stuck with running least squares regressions, and picking out time series that are deemed to be "leading indicators".

Additionally, the influence of academic finance has been damaging. It has come to be expected that investors can "add value" (or "alpha") continuously, at the extreme, on a monthly basis. This makes sense for someone like a market maker. But it creates a bias towards strategies to be long credit, or sell volatility - which means that investors are deliberately exposing themselves to "tail risk".

It is difficult for a top-down macro strategy to look good on such a metric. Economic and financial cycles are long. If you are realistic - that you cannot time the direction of markets each month - you can only hope to capture the bulk of the cycle. (For example, enter equities somewhat above the bottom, and get out somewhere ahead of the top.) But this would mean that you are long equities for years at a time, and performance attribution will say your performance is "just beta".

What To Do About This?


If you are a institutional investor, you have to understand this institutional bias. The typical tactic is to do something else, and just keep top-down macro at the back of your mind. You can have positions that are correlated to your view of the macro cycle, but you cannot bet the farm on macro outcomes. Alternatively, you need to run so many macro positions simultaneously that you get something resembling continuous outperformance (assuming your techniques work, of course).

If you are a retail investor, things are more interesting. You do not have to calculate the "alpha" of your positions, and so you are free to time the macro markets. That said, doing so successfully is a challenge. The most realistic methodology appears to be what I term "passive macro": buying index funds using weightings that you (slowly) adjust based on macro conditions. I doubt that this will ever be an academically-approved exercise, but at the same time, how you choose your asset weightings in a "purely passive" strategy is frankly pseudo-scientific as well ("60-40 worked historically, so it will work in the future!").

If you are interested in the workings of the economy, do not expect true "efficiency" out of the financial markets any time soon. (See my remarks aimed at institutional investors to see why.) Market participants will continue to fuel "rational bubbles" that will periodically blow up the capitalist system. As Hyman Minsky argued, if you want to recreate the pacific conditions of 1950s financial markets, you would need a trauma on the order of the 1930s that acts to reset investor behaviour.

(c) Brian Romanchuk 2014

No comments:

Post a Comment

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.