Goldman Expects Treasury Yields To Surge If Democrats Sweep In November

Over the weekend we reported that in its latest "hot take" on possible election outcomes, Goldman's FX team concluded that a Biden win and Democratic sweep on Nov 3 would "accelerate" dollar weakness, for three reasons:

  • First, Biden’s proposals to raise the US corporate tax rate would make domestic stocks less attractive compared to international markets, all else equal, which could result in Dollar selling if US equities underperform. Regulatory changes, especially anything targeting the technology sector, could have similar effects.

  • Second, a large fiscal stimulus would also likely weaken the Dollar, due to the Fed’s commitment to keep rates low. Normally currencies appreciate after fiscal stimulus, because it lifts rates, and higher rates in turn attract portfolio inflows from abroad. But academic research finds that currencies depreciate after fiscal expansions when unemployment is high and/or central bank policy rates are stuck at their effective lower bound.

  • Third, a more multilateral approach to foreign affairs should reduce risk premium in certain currencies, especially the Chinese Yuan. Goldman recently lowered its 12m target for USD/CNH to 6.50 for this reason: a Biden Administration would likely imply lower trade war risks, and therefore should allow the Chinese currency to gain alongside broad Dollar weakness.

Furthermore, having initially warned that a Biden win would likely be negative for stocks due to the prospect of higher corporate taxes from a Democratic sweep, only to caveat this forecast by saying that a "larger fiscal stimulus and of more predictability in trade policy" could actually push higher, it meant that Goldman still had to opine what Treasurys would do under a Biden administration, and alternatively, what would happen if the "worst-case scenario" for markets, a lengthy contested election struck.

It did so this morning when the bank's rates strategist Praveen Korapaty wrote that "upcoming US elections will likely be a catalyst for interest rate repricing." He first looks at an outcome where the results are contested or delayed for an extended period, in which case Goldman "could see yields decline from current levels, potentially towards previous lows (50bp for10y USTs) in the event that risk sentiment turns substantially negative." Some more details on this scenario:

Delayed results or contested election: One consequence of the pandemic is the fraction of voting occurring via mail is likely to significantly exceed that in prior US elections. Because verification and counting of these mailed ballots does not occur until election dates in some states (including battlegrounds), and given the experience of the primaries, there is a good chance that the result may not be clear on election night. While markets appear to have internalized this possibility, there are additional potential sources of uncertainty. In particular, there appears to be a significant bifurcation in voting method preferences between Democrats and Republicans, with a significant majority of Democrats expressing an interest in voting by mail and a majority of Republicans preferring to vote in person. This could lead to the appearance of a Trump lead on election day, but a potentially significant portion of the Biden vote still to be counted in the mail-in ballots. In a close election, such an outcome could result in claims/counter-claims of victory and/or litigation, and result in significant market volatility over an extended period. While there isn’t an exactly analogous episode in history, we note that it took roughly a month for the results of the 2000 US presidential election to be settled, and over that time, the S&P500 and 10y UST yields declined by roughly 6% and 50bp respectively. Of course, there are some differences between that episode and the current situation to keep in mind. First, economic activity and employment growth were slowing back in November 2000, unlike now, where our economists expect continued recovery. Second, given the proximity of 10y yields to the effective lower bound (of zero), there simply isn’t enough room for 10y yields to rally by a similar amount as in during the Bush vs Gore recount/contested election. In the event of a large decline in equities, we would expect to see 10y yields simply revisit past lows, to the low 50s (in bps). That is, the positive correlation between risk asset performance and yields would be maintained, though with a smaller beta.

Alternatively, in a scenario where there is a clear result, Goldman expects higher US yields. In the case of Democratic sweep of government, yields have the most upside, "largely reflecting the possibility of substantially higher spending." In this case, "we not only expect a large move in the week following the results, but also some follow-through as well, with a cumulative 30-40bp increase 10y yields over the subsequent month." Korapaty also notes that "possible concerns about higher taxes and greater regulation dominate the initial market response, with risk sentiment deteriorating and yields moving lower." In that event, Goldman would urge its clients to go short duration. More below:

Biden victory, Democrats control both House and Senate. Outlines of Vice President Biden’s agenda suggest a significant increase in government spending, only partly financed by higher revenue (through higher taxes, for example). However, given the narrow majority Democrats are likely to have in the Senate in this scenario, it is unlikely the full agenda can be adopted—our economists estimate that the output gap in 2021-22 will be 2-2.5pp lower with unified Democratic control relative to the current baseline. The combination of a growth boost from higher expenditures, a potential COVID vaccine, and an easy monetary stance should all lead to a larger drop in unemployment, but will also likely lead to a wider dispersion and a positive skew to analysts’ forward-looking estimates of growth and inflation outcomes (relative to a divided government scenario). Our fundamentals-based term premium model suggests that the changes on net should lead to upward pressure on bond risk premia (Exhibit 1), to the order of about 30bp. The supply implications of higher budget deficits are also likely to weigh on Treasuries.

To be sure, this scenario could bring forward market expectations for liftoff, and attribution of yield increases could change to reflect increases in both expectations and risk premia. In a recent report, our economists note that FRB/US simulations suggest even a dialed down Biden agenda could result in liftoff occurring in 2023 rather than 2025. If fully priced, we estimate this would result in a 30-50bp increase in the 10y expectations component. However, we do not believe this is likely in the immediate aftermath of the election, especially given the possibility of poor risk sentiment. Instead, we suspect a smaller increase, roughly 10bp or so (similar in magnitude to that seen in the month following Trump’s election in 2016).

Combining the two parts leads Goldman to expect about 40bp repricing of 10y UST yields, which is comparable, though smaller in magnitude than the changes observed following the 2016 election. Going into that election, Clinton was viewed as the status quo candidate, and then-candidate Trump’s agenda was thought to be significantly more expansionary (with a nearly 20pp of GDP increase in debt over a 10-year window). Exhibit 2 above shows the evolution of real and nominal yields (and equities) in the months following that election. As can be seen, much of the increase in yields occurred in the month following the election.

In any case, Goldman would expect the greatest degree of nominal yield curve steepening here—in the month following election results, the bank expects there could be about 30bp of steepening of the 2s10s curve. The real yield curve may prove to be more interesting; while both 5y and 10y real yields may initially increase, increases at the 5y point will reverse over time. The latter is because the Fed may push back against aggressive market repricing at the front end, thereby keeping nominal yields in this sector anchored, even as traded inflation widens.

* * *

Finally, looking at some other potential scenarios, such as a divided government, Goldman expects "a more modest change in yields, given the higher odds of a small fiscal package post-election", to wit:

  • Divided government (Biden Presidency + Republicans Senate or Trump re-election + Democratic Congress). With Republican control of the Senate, we do not foresee any major new fiscal initiatives under a Biden presidency. While there could be some downside risk to sentiment in this scenario (for instance, due to potential roll back of regulatory easing seen under the Trump administration), it could be offset by greater predictability in trade and foreign policy, as well as potentially a more streamlined virus response and public health policy at the federal level. If on the other hand, President Trump were re-elected with Democrats controlling at least one chamber of Congress, we could potentially see a compromise economic package and continued deregulation at the federal level, but persistent uncertainty on trade and foreign policy. On the whole, in either of these scenarios, our economists expect a strong rebound even without additional stimulus, particularly after a vaccine becomes widely available. We don’t see material changes in front end yields in this scenario, given our expectation that there won’t be a significant deviation in Fed policy from our economists’ current baseline. However, we could see some increase in longer maturity yields (and a modest steepening of the curve), given higher odds of a small fiscal package post-election.

  • Unified Republican control. Current public polling1 suggests this scenario is fairly unlikely, as it would require a substantial swing in public preferences for Republicans to recapture the House. Nonetheless, should such a scenario come to pass, we could see some additional fiscal measures and further attempts at streamlining taxes, which may outweigh potential continued trade disputes. In this scenario, we could again see an increase in yields, but not to the extent we would expect in a Democratic sweep.

In conclusion, the chart below summarizes the various scenarios discussed above. As can be seen, all else being equal, except in the case of a contested election (or lengthy delay in result tabulation), Goldman expects yields to rise post-election (which for those familiar with Goldman's recent track record likely means nominal yields will go negative).

The difference between the scenarios is the extent to which the bank expects repricing, with the Democratic sweep scenario being the most impactful. Finally, Korapaty notes that all of these estimates assume there aren’t material changes on other fronts. For instance, if there was a large resurgence of virus cases in late fall or delays in the timeline for a vaccine, the increases we expect in yields may turn out to be more limited.