It is becoming increasingly apparent that the Fed, now data-independent, has just one mandate: keep hiking interest rates until markets break.
That is the conclusion of Citi’s Jeremy Hale, who writes that the FOMC minutes pointed to a view that financial conditions had eased despite the previous Fed hikes because equities had rallied and corporate bond yields declined. Specifically, the minutes note that: “equity prices were high when judged against standard valuation measures.” As a result, the Citi cross-asset strategist believes that “Central Banks are buying in to the BIS thesis that they need to control not just price, but asset price, inflation. In other words, markets may be inflated by easy money.”
In yet other words, those fringe financial blogs you’ve been reading for years were right all along. And, Citi adds, “it also means we might expect Central Banks to remain biased towards tightening until/ unless asset markets turn lower.” Slowly markets are starting to get the message.
What happens then?
To answer that question, Citi has created a hypothetical scenario to assess how various assets may move if the Fed gets its way and takes FF all the way to 3%. According to Citi, this would take real yields to 200bp where they were when the FF rate last held this heady level. Additionally, he expects to see the beta on JPY and EUR real yields as ~20% and 40% respectively. In FX, real yield differentials would likely take $/JPY towards 140 and EUR/$ to 1.05 (and gold to <$800). EM asset markets, which have ignored rising US real yield for now, would not be able to ignore a move that large, and shorting EM $ credit remains the best hedge in Citi’s view.
Some additional details. As Citi starts off, in the past nominal yields rose during each Fed QE: “QE did not lower nominal yields as is the common assumption for reasons that sometimes escape us” Hale points out, underscoring something that has perplexed economists for years. At the same time, breakevens rose strongly, and have been positively correlated with asset purchases. Real yields fell as Fed buying lowered term premia. Overall, nominal yields rose. “So less global QE now, as the Fed winds down, the balance sheet and LSAPs in EA/JP are lower should mean higher real but lower nominal yields” (Figure 2).
So if all we had to worry about was Fed balance sheet shrinkage, Citi’s call would be higher real yields, lower breakevens, lower nominal UST yields. However, that’s not all, as we also have the Fed raising rates. Insofar as this is credible (i.e. yields don’t fall every time they hike), this may also tend to raise real UST yields. Citi’s observations follows:
10y breakevens breached 2% for the first time since 2014 earlier this year while 2016 lows around 1.20% seem secure for now. So they are clearly off the lows. But there is also a general downtrend still evident in recent years. Similarly, real yields fell sharply from the GFC to late 2012 and then bounced to find a new range of roughly zero to +80bp. Current levels around 56bp are still below peaks seen immediately after the US Election last year (Figure 3).
Then there is, of course, risk assets: Hale says that “as our colleague Matt King pointed out, lower flows of Central Bank asset purchases likely also weaken risk asset performance, a view that contrasts with the academic / policymaker view that the stock, not the flow, matters (Figure 4).”
What would happen if real yields continued to rise? What cross market impact should we expect from this?
First we need some idea of where real yields might peak. Our guess, based on Figure 3 is somewhere south of 100bp, probably 80bp or so. But just for a control experiment, lets imagine that the Fed can successfully take the Funds Rate all the way to the 3% currently posited in the long run dots. That probably implies a real Fed Funds rate rise from the current approximately -75bp to ~+100bp. Figure 5 suggests real 10y yields might then rise to around 200bp. Indeed, the last time the real FF rate was over 1% was between May 2005 and December 2007. Real 10y yields ranged from 1.6-2.6% for most of this period, averaging 2.16% (Figure 5).
With that in mind, Citi assumes (i) real yields peaking at 80bp; and (ii) real yields peaking at 200bp presumably over a much longer time frame
What then is the impact on four core asset classes: real rates, FX, gold and risk assets? Hale goes through these one at a time.
The impact on real yields elsewhere obviously depends on drivers. Aside from a hawkish Fed, including possible balance sheet reduction, markets may also expect withdrawal of accommodation elsewhere too. In fact, EUR real yields arguably led USD real yields higher recently following hawkish Draghi commentary (Figure 6 top RHS). But more medium term, EA and US real yields decoupled with the end of US QE in 2014, and the advent of ECB and BoJ asset purchases, so there need not be any automatic rise in foreign yields in response to US ones if the primary driver is a relatively hawkish Fed (Figure 6 top LHS). In fact, the normal rates beta from US to EA and JP is likely lower as Fed balance sheet reduction combines with ECB and BoJ LSAPs, even if at a slower pace than before.
Under our two scenarios of a peak in US real yields of 24bp and 144bp higher respectively, the impact on real yields in Japan and the EA are estimated in Figure 6 (bottom LHS). We estimate a beta of approximately 20% for JP and 40% for EA real swap yields based on three calculations as described in Figure 7.
What about EM? Figure 6 bottom RHS shows that recent improvements in EM fundamentals, the gradual nature of the US real yield rise and ongoing yield grab have allowed nominal EM5y yields to de-couple from the US real yield market. It’s possible this can continue if the real yield move remains limited to recent ranges. But we are not so sure if we end up with the shock scenario of a sharper US real yield rise.
Moderately higher USD real yields are unlikely to be enough to break the recent $ malaise/ range trading. But the more extreme scenario in which markets give the Fed the green light through to a 3% FF rate would likely see EUR/$ back at or beneath the range lows at 1.05 and $/JPY at 140 (Figure 8).
We tend to see gold as a currency more than commodity and one with no Central Bank or yield. Hence when regular Central Bank yields rise, and supplies fall, alternative currencies should do worse. In this light, real yields and gold are inversely correlated as we have shown many times before. Our two real yield scenarios imply gold at $1140 and $752 respectively (Figure 9).
Finally, Risk Assets
As we show in Figure 4, Central Bank plans for liquidity withdrawal would seem to point to significant volatility ahead for risk assets including equity prices and credit spreads.
But we should also be cognisant of recent market experience. In 2015, UST real 10y yields jumped 80bp running into the first Fed hike in December. With a significant lag, equities ended up correcting about 14-15%. In contrast, when real yields rose a similar 80bp in the second half of last year, equities barely noticed (Figure 10 LHS).
One difference is in the data which was weakening sharply in the first episode even before real yields rose (a Fed policy error?). But in the 2016 event, data was strengthening, the global recovery broadening and EPS recovering. In that light the fall in the Citi US data change index currently is interesting. Forget the Citi ESI which always mean reverts over a few months. Here we see the underlying data continuing to show weaker momentum, a trend confirmed by movements in the Atlanta Fed nowcast (Figure 10, RHS). For us, this makes it more likely that the real yield rise, and associated Central Bank confidence, will indeed end up being constrained by market volatility and unpredictable tightening in financial conditions.
Whether or not the market reacts the way Citi predicted is unclear, however it is quite likely that none of this will ever happen: as Hale admits, “we are sceptical about the shift in Central Bank rhetoric given recent data (weak in the US at least) and inflation news (generally downside surprises). Financial stability is often cited as a reason for the shift in policy tone along with weaker arguments about building a buffer for the next recession.”
In other words, the Fed’s attempt to cause a recession as a buffer against the next recession is… unrealistic – and the market knows it, which is also why, despite the Fed’s protests to the contrary and Yellen’s appeals that this time the Fed means to really tighten financial conditions, it won’t be able to do so, without crashing the market, and destroying the wealth effect it spent the past 8 years putting carefully $1 trillion in newly printed dollars at a time. Which is also the bottom line: the market is confident that the Fed won’t casually destroy what it spent half a generation piecing together, with potentially terminal consequences not only for the Fed’s confidence, but for the entire economic system as we know it.
Although who knows, as Albert Edwards said this morning, “something smells different this time“…
Just remember: every Fed tightening usually ends with a financial “event.”