"The yield curve may have lost its predictive power if it no longer provides a signal about the stance of policy"
In recent days the US yield curve has flirted with inversion. As shown in Exhibit 1 from our Guide to the Markets, the two-year Treasury yield is now broadly at the level of the 10-year yield. An inverted yield curve has been a useful indicator of recessions in the past, so this inversion is generating considerable concern among investors.
EXHIBIT 1: THE US YIELD CURVE
Source: (All charts) Bloomberg, Federal Reserve, Thomson Reuters Datastream, J.P. Morgan Asset Management. Periods of "recession" are defined using US National Bureau of Economic Research (NBER) business cycle dates. Past performance is not a reliable indicator of current and future results. Guide to the Markets - UK. Data as of 14 August 2019.
Why has the yield curve proved such a good indicator of downturns? Largely, the yield curve has been a useful predictor of the health of the economy because it contains information about whether the central bank has its foot on the accelerator or the brake. Long-term interest rates should reflect “neutral” policy: i.e. interest rates that do not serve to either boost or restrict activity. A steep curve suggests short-term interest rates are below neutral (and stimulative), while an inverted curve suggests short-term interest rates are above neutral (and thus restrictive).
The time between policy becoming relatively tight and the economy actually slowing has varied greatly in the past, but there is often a lengthy lag. On average, the curve has inverted 17 months ahead of the equity market peak and 22 months ahead of the economic peak. In the mid 1990s, the curve was broadly flat for a period of about five years. Over that period both the economy and the stock market boomed, fuelled by optimism about the promise of new technologies.
It is possible that quantitative easing (QE) has reduced the signalling capacity of the yield curve. QE was designed to purposefully depress long-term bond yields below an economically "neutral" rate. One way of gauging the impact of global QE is to consider term premia—the excess yield investors require to lock up money in a long-term bond. With the slowdown in global growth prompted by trade concerns central banks have committed to further monetary easing. This is having an enormous impact on long-term interest rates and pushing the US term premium to a new low (Exhibit 2).
EXHIBIT 2: US 10-YEAR TREASURY TERM PREMIUM
Source: Bloomberg, J.P. Morgan Asset Management. Adrian Crump & Moench measure of the US 10-year treasury term premium. Data as of 13 August 2019.
For this reason the yield curve may have lost its predictive power if it no longer provides a signal about the stance of policy today relative to neutral. The curve could be flat or even inverted while interest rates still remain stimulative to the real economy and thus are not serving to depress activity.
There could be a causal link between the yield curve and an economic slowdown if a flat or inverted curve depresses activity by reducing the willingness of commercial banks to lend. Net interest margins will be reduced when the curve inverts given that banks typically borrow at short-rates and lend at long-rates.
However, a flattening yield curve has not historically translated into less credit growth, quite the opposite in fact (Exhibit 3). The propensity of the banks to continue lending as the yield curve flattens likely reflects the fact that household and corporate optimism—and therefore the appetite to borrow— tends to build through the cycle. At the same time, impairments tend to be relatively low, which encourages banks to lend.
EXHIBIT 3: US YIELD CURVE AND COMMERCIAL BANK CREDIT GROWTH
%, 10-year Treasury yield minus 2-year Treasury yield (LHS); % change year on year (RHS)
Source: Federal Reserve, Thomson Reuters Datastream, J.P. Morgan Asset Management. Past performance is not a reliable indicator of current and future results. Data as of 14 August 2019.
In summary, we would not rely solely on the predictive power of the yield curve in this cycle. Instead we will focus our efforts on tracking business and consumer sentiment, and the labour market (Exhibit 4), which have all historically provided robust leading indicators of a downturn—and which also provide no obvious reasons to question the signals they are giving this time around. These indicators do show that the US is slowing. At present they do not suggest the US is headed towards a recession but we are concerned about recent developments in the trade war between the US and China and what that will do to corporate sentiment. Indicators of employment intentions are key to understanding whether this turns from a manufacturing led moderate slowdown to a broader economic problem. As a result, we believe a more modest approach to risk is appropriate at this stage, with equity portfolios that have more defensive properties.
EXHIBIT 4: US EQUITY MACRO CORRELATIONS
Source: (Left) Conference Board, Standard & Poor’s, Thomson Reuters Datastream, J.P. Morgan Asset Management. (Right) BLS, Standard & Poor’s, Thomson Reuters Datastream, J.P. Morgan Asset Management. Periods of "recession" are defined using US National Bureau of Economic Research (NBER) business cycle dates. Past performance is not a reliable indicator of current and future results. Guide to the Markets - UK. Data as of 14 August 2019.