Preparing for a Slowdown

The claim:

Bonds are limited in their ability to produce returns that can offset losses on equities during an economic or market downturn. Defensive equity sectors such as healthcare, staples, and certain REITs should have a role to play as alternative portfolio ballasts.

The rationale:

Nominal (as opposed to inflation-linked) bonds clearly have an upper limit on their total returns. Simplistically assume that the yield on Treasuries fell to zero and the price appreciation of a typical bond index would amount to c. 6.5% (duration x yield), excluding any change in credit spreads. The typical move in a recession would be roughly twice that. And with 40% of most portfolios comprised of fixed income, bonds would only produce enough gains to offset an equity decline of c. 4%. That is problematic.

What are the alternatives?

The intuition:

The first point to note is that the equity market is classified by the Conference Board as a leading economic indicator, so declines in the equity market tend to occur well before GDP (spending) begins to fall. A few of the other leading indicators such as housing permits and the ISM New Orders Index (with a few adjustments) have some modest level of predictive power for near-term (e.g. 3 month) returns in the S&P.

References: Standard & Poor’s; St Louis Federal Reserve; Institute of Supply Management; Own Calculations.

The evidence:

Using the bottom quartile of these LEI readings produces the following result, which shows that returns have both a positive skew and reasonable reliability.

For an alternative view, the chart below plots how these sectors perform on a relative basis (vs the S&P 500 Index) from 5 months prior to a recession beginning until 12 months after one has started. The dotted green line shows the median S&P performance in absolute terms across the 8 different recessions since 1969.

This is justifiable from the predictability of earnings. The chart below plots the aggregated sector EBITDA using a representative sample of firms that have been listed since the late 1980s (c. one-third of the current leadership S&P₁). The stability of healthcare (especially pharma) and staples is quite clear, while the volatility of consumer discretionary is impacted mostly by homebuilders and autos.

Scale is logarithmic to show growth rates. References: Own calculations; Standard & Poor’s Index weights.

The timing:

While LEIs (ISM New Orders and Residential Permits) have moderated in recent months, they are not pointing to a contraction in activity yet. So this article is meant as a reference point for later down the road. This is because other indicators with medium term predictive ability suggest that the cycle is more advanced than one might realize.

Notes

1.        As it uses the current composition of the S&P 500, there is an element of survivorship bias which makes many sectors (tech in particular) look better than it is.

2.        Firms are: Tech: Apple, Microsoft, Adobe, Intel, Cisco, Applied Materials, Oracle, Texas Instruments, Cadence Design, Autodesk, Fiserv, AMD, Micron Technology. Staples: P&G, Pepsi, Coca Cola, Walmart, Colgate Palmolive, General Mills. Discretionary: Home Depot, Nike, McDonalds, Starbuck's, Lowe's, Lennar, Target, Ford, TJX, Ross Store's, Dollar General, Carnival, VF Corp. Comm Services: Verizon, Disney, Comcast, Electronic Arts, Omnicom, Interpublic Group. Industrials: Honeywell, Union Pacific, Raytheon, Caterpillar, 3M, Deere, FedEx, CSX, Illinois Tool Works, Eaton. Pharma: JNJ, Pfizer, Eli Lilly, Merck, Bristol Myers. HC Equip: Thermo Fisher Scientific, Abbott Labs, Stryker and Medtronic.

3.        Edited on 19 August to include ‘timing’ section.