Words: 1,450 Time: 6 Minutes
- Goldman sees 10-20% correction over next 12-24 months
- “Big Short” investor – Michael Burry – short sells Palantir (PLTR)
- Better opportunity in the Equal Weight ETF ‘RSP’
The S&P 500 had one of its worst days in ~6 months this week.
It should come as no surprise – the case for stretched market valuations continues to mount. At some point – you should expect the market will pause for air.
Just on this – last week I shared a compelling chart from Robert Schiller…
Using his well-regarded CAPE Ratio – Shiller points out that valuations are as high as we’ve seen in two decades (i.e., the dot com bubble)
And whether the metrics is price to sales, price to earnings, price to cash flow – we are at multi decade highs.
This week there were some other notable bankers sounding the warning bell – the CEO’s of Goldman Sachs and Morgan Stanley.
Goldman’s David Solomon said it’s “likely there’ll be a 10 to 20% drawdown in equity markets sometime in the next 12 to 24 months.”
I think that’s a pretty safe guess…
Morgan Stanley CEO Ted Pick also said: “We should also welcome the possibility that there would be drawdowns, 10 to 15% drawdowns that are not driven by some sort of macro cliff effect.”
My read is both CEO’s are saying expect stocks to mean revert (i.e.. the blue line below)
Where do I think we are in the cycle?
To me it feels somewhere between “c” and “d” (i.e., the FOMO part of the cycle).
I’m reluctant to say we are at “d” – because valuations can easily go higher from here. For example, I’m not sure we have seen “true mania” yet.
But let’s say we are between “c and d” – then buying the market here is more likely to be a poor long term risk / reward bet.
This overlaps with the research I shared from Deutsche Bank’s Jim Reid:
As Reid’s research proves – you want to buy stocks closer to “f” to “g” – where valuations are lower.
Using 100+ years of data across 50+ economies – Reid demonstrated that buying stocks at a lower valuation has a much higher probability of stronger returns over the next 5 to 10 years.
And this is what our two investment bank CEO’s are saying…
Everything always reverts to the mean.
This cycle will not be any different (AI or otherwise). The only unknown is the timing (and why Solomon gave his 10-20% correct a wide birth by saying 12-24 months).
Adam Crisafulli of Vital Knowledge shared his concerns:
“Our biggest complaint about U.S. equities is the extremely disjointed state of breadth, whereby a handful of tech mega-caps have masked some significant red flags beneath the surface”
Yes.
When you have a market this concentrated (where 35% of the market weight is across just 7 stocks) – it’s akin to an inverted pyramid.
And whilst these companies are strong financially (more on this below) – should they not live up to their lofty expectations of growth — the market will likely correct with it.
In other words, the base is not broad enough to stop if falling over.
To illustrate how concentrated the returns have been the past few years – this 3-year chart shows the price appreciation between (a) Mag 7 (pink); (b) S&P 500 (white); and (c) equal weight ETF RSP (orange)
The equal weight ETF assigns a weight of 0.2% to all 500 stocks (regardless of whether it’s “Nvidia” or “Pepsi”- they get the same weight).
This gives us a better sense of market breadth (i.e., the balance of the other 493 stocks)
Equal weight returns have been paltry for 3 years – where all the heavy lifting has been done by just seven names.
But if we look at equal weight vs the S&P 500 over the past 20 years – we find a different story:
Up until the past two years – the equal weight index has continually outperformed the S&P 500.
However, we see the sharp divergence over the past two or so years. Again, look no further than the froth surrounding AI stocks.
From mine, I think there is better opportunity in the other 493 names than the Mag 7.
Case in Point: Palantir (PLTR)
Apart from Nvidia – the other ‘poster child’ for AI has been Palantir.
During the week – their outspoken CEO – Alex Karp – spoke with CNBC following their earnings.
During the interview – Karp had this to say:
CEO Alex Karp ripped into short sellers, calling their moves “market manipulation.”
Karp called the positions “super triggering” and said they are “shorting one of the great businesses of the world.”
“Honestly, I think what is going on here is market manipulation,” Karp said. “We delivered the best results everyone, anyone’s ever seen”
It’s an interesting take.
For those less familiar, Karp was having a direct go at Michael Burry – who has taken large short positions against Palantir and Nvidia.
My first take was short sellers play a valuable role in the market. For example, they help give it much needed liquidity.
But the more important point is I don’t think Burry has an issue with the quality of Panatir’s (or Nvidia’s) earnings or growth.
Again, on the surface it appears Palantir is in the early stages of building a terrific business (it’s still too early to say).
Burry’s view (like mine) is the valuation is too high.
As an aside, I think this is where some investors come undone with high quality companies.
They invest heavily into the narrative or growth story – with little regard to its valuation.
Put another way – buying quality is only half of the equation. You want quality at the right price.
To that end – let me share some quality and valuations metrics for PLTR.
For example, the defense AI company is certainly delivering:
- Revenue – YoY = 38.3%
- Net Income YoY= 127.6%
- Net Income Margin = 32.6%
- Net Operating Cash Flow YoY= 28.2%
- Net Operating Cash Flow Margin = 53.6%
- Free Cash Flow YoY = 27.6%
- Return on Assets = 17.7%
- Return on Equity = 16.6%
And whilst Karp has a lot to be boast about – I’m not sure they are the “best I’ve ever seen” (maybe the Palantir has ever done?) – but they are very good.
I look forward to see how they’re traveling in 5 to 10 years (that will be true test)
But Burry’s central argument (which could be lost on Karp) is what you have to pay to own this stock (at a price of ~$190.74 per share)
- P/E = 596.1 (Price to Earnings)
- P/S = 131.1 (Price to Sales)
- P/FCF = 264.2 (Price to Free Cash Flow)
- EV/EBIT = 790.8 (Enterprise Value / Earnings Before Interest and Tax)
You don’t need a degree in financial analysis to understand how extreme these multiples are.
For example, I consider 15x P/S or 25x EV/EBIT is a very high multiple for a good quality company (e.g., Apple, Google, Amazon etc etc).
But these numbers are simply “off the chart”.
If you’re paying extreme multiples (and Palantir is not alone) – you’re assuming they will continue to post incredibly strong numbers for many years into the future (e.g., cash flow, margins, revenue, return on invested capital etc etc)
And I’m not willing to make that assumption… it’s just not how I choose to invest.
A Top Heavy Market
Andrew Lapthorne of Societe Generale asks two very important questions regarding the AI boom:
- Whether AI can really create profits for companies outside the small group of giants? And second
- Are the funds available for the capital expenditures it will need?
Great questions…
As an aside (and with respect to funding) – I saw Alphabet (aka Google) went to the debt market to raise $26B to fund further AI investment this week. I quote:
- Alphabet Inc.’s $25 billion bond issuance, following similar deals by Meta Platforms Inc. and Oracle Corp., illustrates the reliance on long-term financing to support growing demand for cloud and AI services.
- Alphabet issued $17.5 billion in U.S. dollar bonds and €6.5 billion ($7.48 billion) in euro-denominated notes.
- The proceeds are intended for general corporate purposes, including potential repayment of existing debt, according to Moody’s Ratings.
- The company last tapped debt markets in April, raising €6.75 billion ($7.87 billion) in euro notes.
And whilst the “big guys” are tapping debt markets – they still have a fair amount of cash flow to devote to data centers, but nobody else does (i.e., my inverted pyramid analogy)
John Authers from Bloomberg cites Doug Peta of BCA Research – who offers another way in which AI is “robbing Peter to pay Paul”.
For example, if we consider overall capital expenditure as a percentage of GDP — it is exactly where it was in 2018, before the pandemic.
How is that possible with all the money going into funding this AI boom?
Simple: it has funneled money to a few big players who’ve benefited.
It’s yet to do anything more than that….
Source: Bloomberg Opinion
Putting It All Together
There’s no doubt stocks like Crowdstrike, Palantir, Nvidia and the Mag 7 are posting good numbers.
And there’s every chance that’s likely to continue for a little bit yet.
It’s not the quality of the business I have an issue with – it’s how much we are being asked to pay.
My (investment) thesis is similar to David Solomon and Ted Pick…
That is, expect stocks to mean revert over the next 12 to 24 months (it could be much sooner).
And that correction is likely to be in the realm of 15% to 25%.
As I’ve highlighted in the past – a 25% correction will get us back to a forward multiple of 18x (which is the average multiple over the past 10 years).
If we go back the span of 100 years – that mean is only around 15.5x
That would be a healthy and welcomed development… even if it sees my own portfolio give back some ground.
At that point, have your “quality shopping list” of stocks ready.
Until then – be mindful of how much you’re paying to own the AI names.