Recent Posts

Wednesday, August 28, 2019

Is The Treasury Market Wrong Because Of Hedgers?

Chart: U.S. Treasury and Mortgage Rates

I have some seen some commentary about hedging demand and Treasury yields. One typical interpretation is that hedging demand is pushing Treasury yields "too low," and so the signal from the yield curve is distorted. Furthermore, one could argue that this represents some form of "bond bubble."

Since I do not give unsolicited investment advice to random strangers over the internet, I cannot give an answer to those questions. However, I can safely discuss the background behind these arguments, which have some merit (but are perhaps dangerous).

V-Shaped Yield Moves

The figure at the top of this article shows the 10-year Treasury yield and the Freddie Mac 30-year conventional yield (which I just showed a few weeks earlier, but I cut down the time scale). If you look at the black line showing Treasury yields, you will see that when they hit a local bottom, they tend to violently reverse. I.e., the yield chart is "V-shaped," which means that the price peaks and then falls relatively rapidly.

This sort of peaking behaviour is relatively common in financial markets, but I would suggest that this tendency is magnified in the U.S. Treasury market. (Someone with a good quality database could run statistical tests on that assertion; it did seem to hold for the U.S. market versus other developed government bond markets.)

The usual explanation is that the reversal is driven by mortgage hedging demand. I was never an expert on U.S. mortgage-backed security markets, but the explanation seems entirely reasonable. To summarise, the hedging demand drives yields "too low," and then when rates reverse, the hedging goes the other way -- creating a V-shaped yield chart.

I suspect that this is a factor in the current yield drop. However, I need to add an immediate disclaimer: we do not know where the reversal will happen. For example, knowing that the 10-year yield will rapidly rise 50 basis points in the future does not help directional strategies if you do not know how far the yield will fall before that reversal. (Volatility trades could benefit.) As a result, the existence of hedging "distortions" is not enough to make money trading the direction of rates.

Quick Hedging Demand Primer

The unusual structure of the U.S. mortgage market explains why the U.S. dollar market features these characteristic moves. Conventional mortgages fixed interest rate mortgages in the United States that are financed by the housing agencies offer homeowners the option to repay early. This means that if mortgage rates fall sufficiently, a homeowner can get a new mortgage that pays off the old mortgage, and either keep the same payment and get cash back, or lower the mortgage payment.

In a country like Canada, homeowners do not have the right to prepay the entire mortgage, and so this is not a factor in Canadian fixed income.

Although the ability to prepay mortgages is great for homeowners, it comes at the cost of the owners of the mortgages. There are two main classes of mortgage owners:
  1. levered financial institutions (including banks) that borrow to finance "positions" in mortgages including non-securitised mortgages sitting on lenders' balance sheets);
  2. bond portfolio managers that hold mortgage-backed securities against an index (or actuarial liability).
Since nobody likes seeing entities financing fixed coupon assets with floating rate debt, levered financial institutions typically issue fixed rate debt to finance those assets. This means that all the major holders that matter hold mortgages against some form of liability: debt issues, actuarial liabilities, or a bond benchmark that typically has a large component of non-callable bonds.

Since the mortgages will be prepaid at par -- no matter how low interest rates fall -- their maximum price is capped. However, the fixed rate liabilities keep increasing in value as interest rates fall. This means that they need to somehow increase duration to get their assets back into line with liabilities. There are a wide variety of means of buying duration, but duration itself is fungible, and so all the duration sources move together. In plain English, this means that they are forced to buy Treasury bonds to make up for lack of price appreciation in mortgages. This pushes down yields, creating a positive feedback loop of hedgers driving down yields, which increases hedging demand.

This means that yields fall until they stop falling, at which point the feedback loop turns in the other direction.

Rate Expectations Wrong?

We then run into a interesting controversy: do effects like this mean that "rate expectations is wrong"? (Rate expectations argues that bond yields equal the (geometric) average of expected short-term rates over the lifetime of the bond. If we assume that "expected short rates" are equal to forward rates, this is true by construction. The question is whether "expected short rates" are equal to forward rates, and so we need to interpret "expected rates" as a forecast, and not the mathematical expectations operation.)

I keep running into arguments that "rate expectations is wrong," but in my view, the logic is flimsy. Rate expectations theory accepts the two following points.
  1. If we want to identify "expectations" with "forecasts," the theory would only argue that this is a fair value. Technical factors can move bond yields away from fair value. The question is the durability of these deviations.
  2. There is a term premium as well, and so forecasts are supposed to systematically deviate from observed forward rates. My argument is that the term premium is set by convention, and thus not completely amenable to mathematical modelling techniques. 
In this case, institutional investors are well aware that duration hedging creates a runaway freight train. Nobody will want to jump in front of that train and sell duration until valuation is extremely out of line (i.e., bond yields are stupidly low versus investor forecasts). 

As such, the V-shaped dip in yields is just an overshoot away from fair value. However, these overshoots are quite short-lived, and so it is unclear how meaningful they are.

Explaining Away the Bond Rally?

This discussion raises another interesting issue: the tendency to explain away the bond rally as a purely technical issue.

From both a theoretical and practical standpoint, a great deal of attention is showered on priced market expectations. The argument is that market pricing ought to be a powerful leading indicator.

However, just using market expectations poses a problem for market commentators/forecasters: if we accept them always, everything is always priced-in properly. This makes it impossible to make useful forecasts. As such, forecasters need to disagree with market pricing in order to generate profitable trades and/or attention. The question is whether forecasters as a group go too far in that direction: tending to explain away market moves if they do not fit their priors.

Since I am not positioning myself as a forecaster, I am pretty much forced to take market pricing at face value. This means I will not generate a reputation as a guru any time soon. However, it is an attitude than can be useful in fixed income. The trick is to look at relative value trades, not directional trades. One may hope to identify particular segments of various curves that are mispriced, typically the result of funding flow pressures. Furthermore, one can construct trades with options (conditional curve trades) that pay off on various scenarios, without having much in the way of directional risk. As such, I would argue that fixed income analysts need to hold in check the tendency to want to disagree with market pricing too often.

(c) Brian Romanchuk 2019

No comments:

Post a Comment

Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts.

Note: if you want to post comments from Apple devices (iPhone, iPad), you apparently need to turn off "prevent cross-site tracking" in Safari privacy settings. (The reason presumably is that another URL handles comments, and so the user session needs to be preserved when redirected to that site. I don't like this, but this is not enough to make me switch my hosting service.)