The 10-Year/2-Year Spread: The Most Reliable Recession Predictor?

|
Includes: BIL, DLBL, DLBS, DTUL, DTUS, EDV, GBIL, RISE, SCHO, SHV, SHY, SPTL, SPTS, TBF, TBT, TLH, TLT, TMF, TMV, TTT, TUZ, TYBS, UBT, VGLT, VGSH, VUSTX, ZROZ
by: Long/Short Investments

Summary

The 10-year/2-year US Treasury spread has a general tendency to signal the market’s expectation of an upcoming recession or general downturn in growth.

The 10-year/2-year spread has traded between a 0-250 bp range more than 90% of the time over the past 35 years.

Can the predictable elements of the trend be turned into a viable longer-term trade idea?

Thesis

We are currently in a period where credit expansion in the US continues to exceed nominal GDP growth.

(Source: US Bureau of Economic Analysis; modeled by fred.stlouisfed.org)

The longer this trend persists the greater the probability of a near-term adverse credit event. This generally causes the yield curve to flatten or invert, with shorter-duration securities selling off faster than longer-term securities in anticipation of slowing growth, slowing inflation, and/or an upcoming economic dislocation.

Spreads between the 10-year and 2-year US Treasury generally narrow as the US goes through its business cycle, bottoming during an adverse credit shock, before widening back out during and slightly after a recession.

The trend is cyclical and therefore predictable and could possibly be worth betting on.

Overview

The 10-year/2-year spread refers to the divergence between the 10-year US Treasury bond and the 2-year Treasury note.

In normal economic circumstances, the yield on the 10-year should be greater than the 2-year, creating a positive spread. This signals some combination of positive future growth expectations, positive future inflation expectations, and basic recognition that more adverse economic events are likelier to transpire over a longer timeframe than a shorter timeframe. Thus investors are compensated for taking on the higher risk of longer-duration bonds in the form of higher yields.

The 10-year/2-year spread has gotten recognition for the fact that it has, in a way, correctly identified each of the previous recessions over the past 40 years before they actually occurred by going negative beforehand. (There have been a few false alarms along the way.)

(Source: Federal Reserve Bank of St. Louis)

Expressed differently, a negative 10-year/2-year spread represents an inversion of the yield curve. As abovementioned, this represents the market signaling lower growth and inflation prospects and greater likelihood that a bad credit event is more likely in the near-term (i.e., within two year) than over the longer-run.

Therefore, investors expect higher compensation on near-term Treasuries rather than mid- and long-term durations. This causes the front end of the curve to sell off to a point where yields exceed those on the back end. Those who bet on this event will normally express the thesis using futures contracts, as shorting Treasuries in any material amount can be difficult. (More on this later.)

Based on the chart above, the 10-year/2-year spread looks relatively convincing as an recession prognosticator. But is it genuinely as strong of a forward-looking indicator as it appears?

Both yes and no.

For one, recessions are officially declared after there's already been some substantive indication that conditions are souring. For example, BNP Paribas's infamous "complete evaporation of liquidity" press release on August 9, 2007 finally represented a real-world manifestation that the cracks were starting to show in the US subprime market.

The start of the recession was retroactively applied to the start of Q1 2008. Given we actually had positive growth in Q2 2008 and the official definition of a recession is two consecutive quarters of negative growth (so the term isn't thrown around loosely), the actual official start wasn't declared until Q4 2008's figures came out in February 2009.

But there was no question by late-2007 that there was an issue building in subprime as home prices peaked and started declining, and loans defaults began accelerating. However, the depths of the recession were not yet understood.

The Fed was privy to matters by August 2007, when it began reducing rates from that month forward.

(Source: Federal Reserve of St. Louis)

The 10-year/2-year spread nonetheless went negative beginning in December 2005, as rates rose steadily throughout this period, pushing up the front end of the curve.

We also saw notable drops in 1994 (Mexican peso crisis) and 1998 (Russian crisis), in addition to some choppiness in between with economic tribulations in the Asian Tigers and the downfall of hedge fund Long-Term Capital Management, which was considered the beacon of the asset management industry at one point.

In all these cases, the 10-year/2-year spread's venture into negative digits represented a false alarm as the market disproportionately overestimated the chances of a near-term US recession.

(Source: Federal Reserve of St. Louis)

As we know now, the US equities market began roaring forward throughout the remainder of the 1990's, though because of a massive sector-specific bubble in internet/tech stocks. None of these international events pushed the US into recession despite the yield curve near-inversion (or slight inversion in June 1998). Not until the popping of the dot-com bubble in February 2000 did investors flee and wipe out massive levels of stock market wealth. A recession eventually manifested in Q2 2001.

The yield curve also inverted for Japan's early-1989 debt crisis (i.e., negative spread), which dragged the US into negative-growth territory by Q3 1990.

(Source: Federal Reserve of St. Louis)

We also notice that during each recession we see a widening of the spread to indicate bettering economic conditions as inefficiencies are shed and positive growth resumes.

Where are we currently?

So this naturally brings up the question of where this metric is presently. At the moment we're at a spread of 122 basis points. If we look at the long-term chart, 250 bps serves as a sort of "resistance level." The 10-year/2-year spread has never materially ventured too far beyond that rate.

(Source: Federal Reserve of St. Louis)

Also, negative spreads generally aren't sustainable in the long-run either. While credit shocks in the economy will continue to happen, a 1930's type of extended depression is unlikely with the monetary tools modern-day policymakers have at their disposal. A standard shaped yield curve is highly damaging to the banking sector if maintained for longer durations of time. As the banking system is the conduit by which monetary policy flows, it is in central bankers interest to avoid this.

So essentially we're right in the middle of this generalized 0.00%-2.50% range.

In terms of future direction, the trend is down. As economies expand throughout the business cycle, credit growth generally increases at a rate faster than nominal GDP growth, which increases risk in the system. Accordingly, chances of a negative credit event in the near-term increase, growth/inflation expectations decline, and the curve begins to flatten out, compressing the spread.

The spread was as low as 76 bps last summer, but has since held in the 114-134 bp range shortly after the November US elections, with the spreads widening back out from expected pro-business policies from the Trump administration that would boost future growth and inflation rates.

However, as stated, with credit growth continuing to exceed nominal GDP growth, it's still expected that this spread will continue to trend down over time. Fed tightening also helps narrow the spread, just as loosening helps widen it and is a big reason why we see a big jump in the spread during recessions when monetary accommodation is a common policy tactic.

Trading this idea

Going to the futures market is one option, by shorting contracts of the 2-year and going long contracts on the 10-year.

Another option is less specific to the 2- and 10-year Treasuries, respectively, but supports the same general thesis.

One can short a low-duration Treasury ETF and go long a high-duration Treasury ETF. TLT (NYSEARCA:TLT) has an effective duration of 17.23 as of this writing and yield of 2.59%. SHY (NYSEARCA:SHY), a 1-3 year Treasury bond fund, has an effective duration of 1.93 and yield of 0.74%.

Betting on a continued narrowing of this spread would entail going long TLT and short SHY. This is essentially the same thing as going short the front-end and long the back-end of the yield curve. Even when including brokerage fees and ETF-specific expenses, this will generate positive carry of about 1.5% (i.e., essentially a "net dividend").

Best of all, no capital outlay is required as the short generates a credit, which can effectively be applied to the long side of the trade. Therefore, without net capital outlay, one doesn't have to be totally time sensitive to the idea working out if the type of return isn't meeting annualized gain targets.

This is one idea that I think has the potential to squeeze a little more return out of a portfolio over time, but is more of a longer-term idea.

Main Catalysts

- Credit growth continuing to exceed nominal GDP growth, which gradually increases the probability of a near-term credit-related shock. This would expect to cause shorter-duration securities to sell off at a faster rate than longer-duration securities.

- Federal Reserve continuing to hike rates, which tends to narrow spreads.

Main Risks

- Continuation of accommodative monetary policy - Fed continues to remain behind the curve.

- US economic sentiment continues to improve such that near-term credit-related risks remain muted.

- General mistiming of the US business cycle.

Disclosure: I am/we are long TLT.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: Short SHY