Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
US stock valuations have been in nosebleed territory. But this doesn’t mean valuations can’t increase. And according to the hedge fund Hudson Bay Capital, this is probably what they will do. Why? Because MAGA, basically.
It should probably be noted up top that Hudson Bay’s research side is mostly known for publishing Stephen Miran’s infamous User’s Guide to Restructuring the Global Trading System. Miran is now Trump’s man on the Fed board, so the hedge fund’s arguments are worth examining.
In a paper titled No, Stocks Aren’t in a Valuation Bubble — Not Even Close, Hudson Bay’s strategist Jason Cuttler now argues that recency bias, anchoring bias and “myopia” mean that investors risk missing out on “generational upside”.
What does generational upside even mean? In what reads as a full-throated defence of listed stocks, the “ultra-optimistic fair value” that Hudson Bay consider “plausible” for the S&P 500 is 25,000 — so up around 280 per cent. And their time horizon? “Today”. Private equity can put that in their IRR model and smoke it.
To be fair, some of the 27 scenarios they forecast are less optimistic. Some even see market falls. We’ve thrown them all into a chart:
As Hudson Bay’s Cuttler writes, “these levels present an extremely wide range”. But:
. . . they also highlight the inherent option value of being long equites [sic], which cannot go below zero (-100%) but could plausibly reach multipes [sic again] of their current level ($25,000 is 3.7-times the current $6700 index level).
It’s always reassuring to be reminded that you can’t lose more than all of your money. And it’s interesting to see this evidential fact turned into a bull case for buying stocks.
But how did Hudson Bay arrive at its forecasts? By deploying what it calls a “thoughtful historic regime analysis”. Let’s take a peek inside.
The hedge fund starts with the principle that high growth stocks (eg Apple) should trade at higher multiples than low growth stocks (eg Alcoa). Fine. It then reckons that lower long-term risk free rates should mean higher stock prices, on the principle that a stock price is just the present value of future cash flows. Again, no kidding.
But what Hudson Bay then does is subtract the earnings yield (aka the reciprocal of the price/earnings ratio) from the US Treasury bond yield and call this difference the “sentiment spread”.
Market pros will recognise this as a rebrand of what is usually referred to as the Fed model residual. But it’s a twenty page repackaging. And if Hudson Bay can assign the right value to this residual/ “sentiment spread” it’s got the secret sauce to accurately calling the market and making a tonne of money. So it’s worth our time. The hedge fund’s answer? Vibes.
In the dark days of Nixon, Ford, Carter, and *checks notes* Reagan’s first term, the earnings yield exceeded long bond yields. Hudson Bay calculates the median difference during what it calls an era of pessimism at -0.3 per cent. And the median “sentiment spread” fell to a whopping -3 per cent during the “stagnation” period of 2007-2023.
But during the go-go years that kicked off with the launch of Microsoft Windows and ended with 9/11 terror attacks the median “sentiment spread” was 1.8 per cent.
We’re not sure where Hudson Bay sourced its earnings yield data, but we grabbed data from Robert Shiller’s site to give you a sense as to how this all looks in a chart:
How should we characterise the current era? Hudson Bay reckon that:
. . . an “optimism regime” would likely be achieved through a revival of US manufacturing, rising domestic labour market participation, AI innovation percolating through the economy, and a stable world order under US hegemony. Such outcomes are consistent with President Trump’s “MAGA” policy goals which include higher tariffs to motivate domestic manufacturing, restricting immigration to reduce supply of foreign-born workers, rolling back regulation to allow AI and other new industries to thrive, and a US-led world order.
Readers may have differing views as to whether these goals are consistent with an optimism regime. But that’s what makes a market.
We rebuilt their framework to show how different “sentiment spread” assumptions are consistent with different fair values for the S&P 500. And rather than limiting ourselves to the mere 27 scenarios set out by Hudson Bay, we applied it to 374 scenarios for each of their four earnings per share assumptions, just so you can see how the numbers roll. Hover your pointer over a circle to see the S&P 500 index value associated with each bond yield and sentiment spread value.
The dotted line shooting up from the horizontal axis shows today’s values for the long-term bond yield. The one darting to the right of the vertical axis shows our calculation of Hudson Bay’s “sentiment spread”.
You’ll notice that the “sentiment spreads” are different for each of the earnings scenarios, reflecting the fact that discounting lower earnings with a low risk discount rate can get you to the same present value as discounting higher earnings with a high discount rate. If that sounds like financebabble, just think 1/10 = 10/100: this is pretty much the point.
And so you can see that you can get to any S&P 500 fair value number you want using this approach — depending on where you think bonds will trade and how thrilled everyone expresses themselves to be living in MAGA world. Alphaville has coloured in red the combinations of long bond yield and sentiment spreads consistent with lower market valuation and in blue the combination consistent with a higher valuation.
If you think we should be Reagan II-Clinton-era thrilled, go back up to the dataviz and hover your mouse over the circle in the consensus box where current bond yields and a “sentiment spread” of 1.75 per cent coincides you can see that an S&P 500 fair value of around 9,280. And all we need to do to get there is all believe.
Alphaville thinks it’s great that Hudson Bay is stretching the historical analysis window back to 1971. Even if we’re suspicious that using their framework tells us anything about whether stocks are over-, under- or fairly valued, we agree that recency bias is a dangerous thing.
But we wondered what the data would look like if we stretched back even further:
Elongating the time window from fifty-five years to one hundred and fifty-five years, it turns out that the average “sentiment spread” was -2.6 per cent. Aside from the 1985-2000 “optimism” era that Hudson Bay believes that MAGA could reproduce, there are no other prolonged periods in which bond yields were higher than earnings yields.
To be clear, we’re not saying that the Fed model residual should be mean-reverting over pretty much any period — short or long. But if you were so minded to plug this average number into the Hudson Bay framework using consensus earnings, S&P 500 fair value falls to 3,809.
So thank goodness we live in a brave new world where things like discount rates and future cash flows have lost all meaning.
www.ft.com
#triple
