Let’s See If We Can Find A Reason Why The Plunging Dollar Isn’t Trump’s Fault

Ok, look…

We’ve gotten to a place now where we’re all searching for reasons why we shouldn’t blame Donald Trump for adverse market outcomes.

In other words, the default when looking at things that are directly influenced by policy decisions (think FX and yields), is to assume that the only data point which matters is the fact that there’s an increasingly unhinged reality TV show host whose CV includes a WrestleMania cameo, jet-setting around the country, launch codes in tow, holding delirious campaign rallies nine months after the election and dog-whistling to Klansmen.

Here’s what we wrote earlier today:

Headed into 2017, “long USD” was one of the consensus, “no-brainer” trades. Hilariously, the excuse everyone would have given you back in late December for their rampant bullishness is the very same excuse they’ll give you now to explain why they’re overwhelmingly bearish: “dude, have you seen who’s President?”

Let’s just be honest with ourselves here: there is no question why yields and the dollar have had a harder time launching this year than one of Kim’s early-model, homemade ICBMs. The culprit is Trump.

Now if you want to just go ahead and start from that, and then move on to talking about how Trump is affecting the rate diffs pillar for the dollar and how he’s now a liability to his own agenda which in turn is making the Fed nervous about whether fiscal policy is prepared to take the baton from monetary policy, then fine.

But do me a favor: don’t act like the controlling factor here isn’t Pennywise, the geriatric orange clown.

So with that as the setup, consider the following from Cameron Crise (or, as Jeff Gundlach knows him, “who?!”)….

Via Bloomberg

Everybody loves a good narrative, from sportscasters to investors and market pundits. “Joe was clutch because he wanted it more” is standard fare among sports announcers, just like “x happened because of y” is a common post hoc narrative attribution for market observers. “The dollar is weakening because of Trump and U.S. political dysfunction” is a common talking point at the moment, but is it true? The evidence suggests that there may be a “Trump effect” on the dollar, but it’s probably smaller than you think.

• At the start of 2017 the bullish backdrop for the dollar seemed obvious. Not only was growth solid and the Federal Reserve tightening policy, but the Republican clean sweep in the elections raised the prospect of meaningful fiscal and regulatory reform.
• Since then, the Fed has delivered the anticipated tightening while the government has whiffed on meaningful policy shifts while descending into scandal. So clearly the dollar’s weakness is attributable to the latter, right?
• Well, not necessarily. After all, U.S. inflation has consistently disappointed, and while the Fed has indeed remained on track for its 2017 policy agenda expectations for future tightening have been scaled back dramatically. At the same time, conditions elsewhere in the world have generally improved, both economically and (in the euro zone, at least) politically. This shift in fortunes seems to at least partially explains the weakness of the dollar.
• This type of analysis still reflects a narrative approach, however. What do the numbers say about the influence of politics? It’s difficult to isolate politics as a explanatory factor for financial asset pricing, but presidential approval ratings from the American Presidency Project seem a reasonable place to start.
• Earlier this year French presidential polling was a clear driver of pricing in EUR/USD. Over the past few months, however, the pair has overshot the model that worked very well through May. Introducing Trump’s approval rating into the regression appears to dramatically improve the performance: the model output rises from 1.10 to 1.16.
• So that’s it, then? Trump’s been worth 5%-6% to the dollar? Not necessarily. On a daily return basis adding in the Trump factor barely improves the model’s performance. Moreover, there’s essentially no improvement to a USD/JPY model when introducing Trump as an explanatory variable.
• Over longer periods of time, presidential approval does show up as a statistically significant input into FX return models. However, the degree of improvement over a more basic model is small, and the magnitude of the influence is fairly inconsequential. In both EUR/USD and USD/JPY, a 10% swing in presidential improvement implies a 0.40% change in the dollar over a six month period.
• Trump’s current influence is probably larger than that, but almost certainly not as much as today’s narrative would have you believe.