The Case For Owning Treasuries Is Evaporating
Authored by Simon White, Bloomberg macro strategist,
Contrarians will love this column as it will show there is still no smoking gun for a recession in the US.
They might see that as proof then that one must be imminent and yields are going much lower. But contrarianism for its own sake is rarely a good investment strategy.
With low recession risk, there is no need to own Treasuries for those that don’t have to, and with increasingly entrenched inflation, there is little reason not to short them.
Investors have been hesitant to short bonds, but the diminishing risk of a near-term NBER-defined downturn and Wednesday’s stronger-than-expected CPI data, the fourth in a row, should squash any lingering doubts.
No imminent recession removes (for now) the need to own bonds for protection.
Yields look too low, especially as rising inflation risks mean investors will become increasingly reluctant to fund the US government without a bigger margin of safety.
Further, the yield curve should begin to steepen again. As the chart shows, in either a hard or soft/no-landing scenario, the curve has historically started to steepen at this stage after the last Fed hike. (More rate hikes will likely be on the table at some point, although not until after the election.)
No near-term recession also removes (again, for now) typically the single-biggest risk that stocks face. The current bull market, perhaps surprisingly, is only performing in line with the average historical bull market. That means (despite some near-term risks, e.g. from momentum) the bull trend should remain intact. As Wednesday’s CPI data showed us though, the path is now likely to be bumpier as the market adjusts to the potential of higher real yields.
An agnostic, data-led approach to markets is best. That showed in the fourth quarter of last year that an NBER recession was less likely than so over the next 3-6 months (reversing my call from earlier that year that a near-term recession was on the cards). Based on this framework, a recession remains unlikely over the coming 3-6 months.
NBER recession dating is not an exact science, but the research body identifies the four key variables it uses in its assessment:
Industrial production
Payrolls
Real personal income expenditure (PCE)
Real personal income net of transfer payments
In every post-1970 recession all four of these were contracting on an annual basis (apart from 2001 where three out of the four were contracting). Currently three of the four are still expanding at about the same rate they were six months ago, while industrial production is essentially flat, also unchanged from last year.
NBER recession-dating happens often long after the fact. On top of that, the four indicators above are all coincident-to-lagging. But leading indicators show that each one is projected to remain supported over the next six months. (One should never base a view on only one leading index, but for each case there is more than one supportive indicator. For brevity though, I’ve only shown one for each here.)
The US leading indicator expects industrial production to pick up.
The fall in claims anticipates more payroll growth.
Easing bank credit should help retail sales and personal consumption stay supported.
And Fed wage survey indicators expect steady wage growth, which is currently positive in real terms.
There are bones one could pick with this. The jobs market overall is showing some potential signs of weakening, such as the rise in part-time employment and the fall in the quits rate. Household employment is diverging negatively from payrolls, and the number of US states with a rising unemployment rate is increasing, But the effects of unaccounted for immigration are potentially biasing the household survey weaker, and the strength of the labor market probably lies somewhere in between it and the payroll survey.
There are other data points that may give pause for concern, such as the rise in loan delinquencies due to unemployment, or some of the details in the latest NFIB small business survey. But picking and choosing data points to support a recession thesis without any systematic process that incorporates the strength of their relationship to recessions, by how much they lead them by, and how noisy they are, is not a robust one.
The data in totality is not currently supportive of a near-term NBER recession. Furthermore, rapid downward revisions in the four key data points are less likely (although not impossible) when leading data - chosen such that it is minimally revised and therefore itself not subject to future sizable changes - is turning up as it is today.
Bond positioning - based on the bond future proxy in the chart below - likely got a lot longer late last year when perceived recession risk was rising (even though, as discussed above, the risk was low). But if a recession continues to look off the cards for the next 3-6 months, there is less reason for multi-asset managers and their like to own bonds.
That’s even more the case now that it’s evident inflation is not going anywhere soon. Moreover, leading indicators, such as the easing in bank credit, show inflation should soon start rising again.
Owning bonds, unless you have to or in an extreme geopolitical-driven flight to safety situation, is an increasingly suboptimal proposition in this environment, while shorting them looks more and more attractive.
There will be dip buyers now that yields have hit 4.5% (although in a sign the wind may be changing, demand at Wednesday’s 10-year auction was terrible), but that level will likely prove to be low with nominal GDP rising at an annual pace of 5.9%. The contrarian trade is still for higher yields, but probably not for much longer.