Overall, the analysis moderately increases our confidence that today’s economy should be able to weather fairly large rate increases, without spiraling into a recession.
I’m sure Goldman didn’t mean for that to be funny, but it damn sure is.
It’s like saying: “I thought about it and overall, it seems like my house should be able to weather a fairly large storm without collapsing on top of me and my family.”
It’s comforting. Sort of. But sort of not. Because you know, what’s “fairly large”? And also, why are you thinking about that in the first place? Is that something you think might be relevant soon? And finally, what’s with the hyperbole there at the end? “Spiraling into a recession”? Was that really necessary? And if so, why?
Humor aside, it makes sense in context. Goldman is out stress testing the economy for a rates shock and part and parcel of a “stress test” is positing an extreme (or relatively extreme) scenario.
The note reads a bit like a companion piece to something they did in January on the effect a “Black Monday” redux would have on the economy. That analysis centered around the the impact a sudden sharp decline in equity prices would have on financial conditions and the knock-on effect that tightening would have on the real economy. This quote from the January note captures the gist of it pretty well:
We have argued that the most important reason for the acceleration in growth last year and for growth optimism in 2018 is the sharp positive swing in the impulse from financial conditions. The run-up in the equity component of the FCI has accounted for roughly half of the 137bp index easing in 2017 and 80% of the of 32bp easing year-to-date.
The bottom line in the stress test piece (the new note) is that a sharp rise in rates would tighten financial conditions by 200bp and of that 200bp, 110bp of it would be attributable to a 20-25% decline in equity prices.
This is a pretty straightforward exercise. One thing they note right off the bat is that up until this month, rate rise was well digested because it was down to positive growth expectations. “It is only in February that expectations of more hawkish policy and higher inflation have also significantly driven the rates sell-off,” Goldman writes, adding that “while activity tends to decelerate if the increase in rates is driven by more hawkish monetary policy, higher inflation, or wider risk premia, it tends to accelerate if the increase in rates is driven by positive growth news.” Here’s the chart:
The bank goes on to note the obvious which is that when it comes to financial conditions, rising stock prices, the weaker dollar, and tighter credit spreads have trumped rising rates and therefore, the read-through for growth (the GDP impulse from Goldman’s FCI) is still on the upside (and materially so).
So what happens in the event we get a 2-standard deviation move in 10Y yields that takes us all the way up to 4.5% by the end of the year? Well, Goldman uses the FRB/US model to estimate that and the results are as follows (the “baseline” here is just Goldman’s 3.25% forecast on 10s):
Exhibit 4 shows the growth rates of GDP, consumption, business investment, and residential fixed investment in the baseline and stress scenarios. In the stress scenario, year-on-year GDP growth slows to 2.25% in 2018 and 1% in 2019, well below our forecast of 2.8% and 1.9% respectively. The model implies peak drags on GDP and consumption growth of 1pp and 0.5pp respectively, while the peak hit to business investment growth is 2.25pp. Higher interest rates weigh the most on residential fixed investment with a peak hit to growth of 6pp, pushing the sector in 2019 into a short “housing recession”.
Ok, so there’s that, but the headline grabber from this analysis comes when Goldman says that because risk assets are prone to doing crazy shit (they don’t phrase it that way), the model-implied 20-25% hit to stock prices could well prove to be conservative.
Here are the two charts that illustrate the point above about what the FRB/US model spits out for equity prices and Goldman’s FCI in a rate shock scenario:
“We think that volatile risk asset prices may initially respond more to the rates shock than implied by the fundamentals-based FRB/US model,” Goldman goes on to suggest.
That’s probably true, but do note the use of the term “initially.” The impact of plunging stock prices on the real economy probably operates with a lag. If the move is fleeting then it’s less likely to have a pronounced read-through. Take this month’s correction for example. This was a technical selloff exacerbated by the rebalance risk in VIX ETPs being realized and then triggering a cascade of forced de-risking by a systematic crowd that was running (in some cases) record high equity exposure going in.
Once that abated, the storm passed, probably before most of Main Street even knew what was going on (although based on SPY flows, some of the folks who opened all those new E*Trade accounts in January were getting the fuck out of dodge).
So I mean, if what you want to do is posit a worst case scenario here you definitely can, but you should note that Goldman didn’t set out to imply that because risk assets would probably overreact initially to a 2-standard deviation move higher in yields that what comes next is armageddon. Rather, they were simply stating the obvious which is that given the market’s hyper-focus on yields, 4.5% on 10s would probably lead some people (and some algos) to do irrational things.
Anyway, the rest of the piece is upbeat, with Goldman noting that while an FCI overshoot (i.e. a severe tightening due to an equities freakout) is certainly possible, a rise in rates of the magnitude posited in the stress test would likely entail at least a certain amount of further optimism around the growth outlook. Additionally, they remind you that the bulk of outstanding liabilities are fixed rate and also that because debt servicing costs are so low, there’s probably considerable room for them to rise before the proverbial shit hits the fan.