At ARM’s Length

Recent releases of CPI and PPI are genuine news, not only because inflation is a concern of every American, but also because they contradict the prevailing narrative that it’s being brought under control. But most of the financial media has allowed news of an initial public offering (IPO) to eclipse it in scarce headline real estate, even traditionally wary and skeptical outlets.

Make no mistake … artificial intelligence is real … and it’s spectacular. It will change the world. It’s the excessive hype that burns me up. AI has been around for years, and it will be around for many more. The new thing here isn’t the technology, it’s the buzz. It’s a concerted effort to gin up excitement in order to sell the public securities at inflated prices, encouraging emotion to discourage sober number crunching.

Those of us who have been around for a while are getting big time deja vu.

By number crunching, strictly speaking I mean reading financial statements and performing discounted cash flow analysis to estimate a reasonable price to realize the return you seek given the risk. You know, the kind of stuff that made Warren Buffett rich. You can be sure the nice folks selling the investment have thought it through. Unfortunately, most “retail” (their respectful word for you) investors can’t be bothered … it’s boring and easily drowned out by dazzling stories that engage the emotional centers of the brain. And the financial media know it.

”It’s ten times oversubscribed!” (psst … hey buddy, everybody wants this … don’t miss out!)…

That’s one of the reasons I so like this quote from Sun Microsystems’ CEO Scott McNealy. It packages the concept of number crunching in a brisk, digestible form accessible to even the most innumerate investor:

”At 10 times revenues, to give you a 10-year payback, I have to pay you 100% of revenues for 10 straight years in dividends. That assumes I can get that by my shareholders. That assumes I have zero cost of goods sold, which is very hard for a computer company. That assumes zero expenses, which is really hard with 39,000 employees. That assumes I pay no taxes, which is very hard. And that assumes you pay no taxes on your dividends, which is kind of illegal. And that assumes with zero R&D for the next 10 years, I can maintain the current revenue run rate. Now, having done that, would any of you like to buy my stock at $64? Do you realize how ridiculous those basic assumptions are? You don’t need any transparency. You don’t need any footnotes. What were you thinking?”

Another great thing is the context … the aftermath of the dotcom bubble. This is where the deja vu comes in … investors were sucked in with exciting visions of the future and overlooked crunching the numbers. In the cold light of day, after they’ve lost their retirements, investors blinded by emotion will one day once again be asked “What were you thinking?”

The correct answer, of course, is ‘I wasn’t. I was feeling.’

As a rule of thumb, a shorthand way to screen out the worst offenders is to watch for hype. If it’s something you can buy and it’s all over the headlines, your spidey senses should go on alert. They’re not trying to make you money; they’re trying to get your money. After thirty years of investing, the more exciting the pitch is, the less I want it. If you’re speculatively inclined and calmly, rationally anticipate a greater tool to soon pay you even more for it, it might be something you do with money you can afford to lose.

But keep it at ARM’s length.

11 thoughts on “At ARM’s Length

  1. Milton Kuo says:

    This talk about being wary of all the AI hype was the subject of an article in Institutional Investor. I believe some of the ugly dynamics of venture capital investing were discussed on iTulip some years ago but they’re in full evidence in this bubble cycle. First the dreadful cryptocurrency bubble and it seems VCs have now shifted toward the AI bubble. One of the worst actors in all of this garbage is Andreesen-Horowitz.

    1. Finster says:

      Thanks Milton. To use the McNealy analysis for context, I tried to find price to sales figures for ARM. Given all the coverage it’s getting, you would think it would be easy. It’s probably out there somewhere but buried under a haystack of narrative. Price to earnings is well north of 100 times. Any buyer at these prices has to wonder whether the seller is the smart money.

      I know very little about Arm so wouldn’t presume to discuss details. I have no beef with Nvidia as a business. Its tensor processors are a marvel of engineering, regardless of use case, and appear to have a bright future. But the shares are priced for the hereafter, also well north of 100 times PE. 2750 times dividends. 2750 years to get your purchase price back is kind of a long wait for most human investors. At least the ones who aren’t counting on selling to a greater fool.

      In all fairness, those dividends are likely to grow, but you still have to be pretty darn confident nothing will go wrong for the decades it would take to get a reasonable return. No competitive threats, no yet newer technology, the world pretty much stays the same while this growth catches up to today’s share price.

      I like the company, but would be hard pressed to justify the shares at a tenth of where they’re currently trading. I haven’t analyzed them either, but it’s very likely shares of the company that actually makes NVDA’s processor designs, TSMC, or the company that makes the equipment used to make the processors, ASML, would be better values.

      In the 1990s, Intel, the company whose processors powered the personal computing revolution, appeared poised to take over the world. By June of 2000, INTC peaked at around $75. A little more than two decades later, the shares are trading around half that level.

      1. Milton Kuo says:

        NVIDIA is a real company and it is well-managed enough that it has won the video card wars. You can go back in time and look at all the companies that designed and manufactured video cards and, at least to me, it was far from obvious that NVIDIA would be the company that ultimately out-survives everybody else. I can remember companies like Diamond Multimedia (been acquired and now made non-video card things), ATI (I thought they would win but ultimately got acquired by AMD), Matrox (still around but tiny and offering very niche products), 3dfx (they were huge in the late 1990s but somehow screwed up and got acquired by NVIDIA), and a few other companies that were probably repackaging some other company’s designs. As of today, the major players are NVIDIA and AMD (formerly ATI and their products are not as good as NVIDIA’s). I think Intel is trying to made graphics processors but they’re not even close to the being in the same league as NVIDIA and AMD.

        The point being, NVIDIA is a fairly mature company (easily over 25 years old) and, in my opinion, it really hasn’t developed an entirely new business that is disrupting things. For its entire existence, it’s been graphics processors or video card. NVIDIA seems to be the “luckiest” company I’ve ever seen in that it first got a tremendous boost in sales thanks to the cryptocurrency bubble (video cards using NVIDIA’s GPUs were highly sought after to perform the computation for “mining” cryptocoins such as Ethereum) and now that the cryptocurrency bubble is deflating a huge AI bubble has been created that juices sales for NVIDIA’s GPUs. Either either buy NVIDIA product or somehow be large enough to design one’s own silicon. I’m not sure if those companies designing their own silicon (such as Google and Amazon) sell their product to others.

        NVIDIA is so overpriced as a stock that I actually shorted it very early this year and have taken a beating. 🙂 It is even more overpriced now, in my opinion. At the time I shorted the stock, I felt that the price would fall about 90%. I think this whole AI nonsense is not going to be real and it’ll ultimately crash spectacularly. This means that I think that at current prices, NVIDIA’s stock will fall even more than 90%. In the meantime, however, the company does have a real product and it is making real revenues and profits (extremely high margins on their $40,000 top-of-the-line processor). I just don’t think those revenues and profits are anywhere near sustainable.

        It’s been a while since I looked but I think NVIDIA’s stock price was selling at something like 40 times trailing 12 months sales. 😀 Sun Microsystems’ 10 times sales is a piker compared to NVIDIA. Both Sun and NVIDIA at their peaks (and this is assuming we’ve seen a peak in NVIDIA stock) were mature companies. Sun’s stock price ultimately fully reverted to mean and the company was ultimately ground by less expensive Intel-based servers. It’s not obvious if there is a competitor that will take away NVIDIA’s business (I think “no” on the GPU front; “maybe” on the specialized GPUs used in AI) but I think the demand for AI processors will fall precipitously and, at $40,000 per processor, those kinds of margins are going to attract competitors if sales hold up.

        As for ARM, it’s an old company that was founded in the late 1980s. Intel actually owned IP and I believe that allowed them to make ARM processors without having to pay ARM royalties. The Intel product was called StrongARM and I actually did some work with it in the early 2000s. You mention Intel’s stock and it’s down so far from its dot-com peak due to firstly a bubble and secondly due to poor management. You probably won’t be surprised to know that Intel sold off StrongARM to focus on Atom, the low cost, crappy X86 architecture chip.

        Intel could have been one of the suppliers, either as a fabrication facility or selling their own StrongARM processors, to the billions of devices that use ARM processors today. As it is, it sells nothing and the Atom processors, to the best of my knowledge, are not very profitable and don’t sell in nearly the numbers of ARM. For what it’s worth, I sold Intel stock in the high $60s during the dot-com bubble and never looked back.

        ARM stock is clearly a bubble and I don’t think sales will ever grow enough to justify its current valuation. It will probably try a combination of increasing the licensing fees it charges and organic sales growth. It may or may not be able to raise prices too much due to the existence of RISC V which doesn’t have licensing fees. I believe China is making its own processors based on RISC V designs as it seeks to restructure its economy to be less dependent on foreign technology so that is another way RISC V could prove to become a competitor to ARM in the future.

        I saw some numbers on the ARM IPO and I think P/S was over 10 (LOL) and the P/E was insane as you stated. SoftBank is probably the dumbest institutional investor on the planet and they’re very lucky that they’re currently sitting on a rather small gain on their ARM investment. My gut feeling is that when this is all over, SoftBank will end up taking a huge loss on ARM.

        Here are two articles, one short and one very long, about the history of ARM. Interesting reading and it’s amazing how quickly ARM proliferated with the introduction of the iPhone.

        1. Finster says:

          Thanks again, MK … those links add some good historical perspective too.

  2. jk says:

    masayoshi son sold shares back and forth between softbank and his vision fund at self-created prices to pump up arm’s capitalization.
    doomberg summarized the transaction as follows:
    “With the financial world anxiously awaiting the start of Arm’s roadshow in the hopes that Son can reignite the flaccid IPO market, Son has decided that his chip-maker is worth $64 billion because Son agreed to buy a chunk of it from Son for that price. In so doing, Son agreed to pay Son twice what Son sold it to Son for six years earlier.”
    some valuation details from the same source:
    “At Son’s self-declared valuation of $64 billion, Arm would sport a price-to-sales (P/S) ratio of 25 and price-to-earnings (P/E) of over 120—nosebleed numbers in any market environment. …
    Perhaps more concerning is the litany of material risks disclosed, in particular, those surrounding its deep dependence on China. Arm only owns 48% of Arm China, and the company is greatly exposed to the whims of the Chinese government.”
    arm china is the source of 24.5% of arm’s revenue, and has a history of late payments to arm itself.

    1. Finster says:

      Wow JK … so maybe it’s not the “smart money” selling, but the &!#+€£$ money selling…

      1. Milton Kuo says:

        It’s only been trading days but ARM’s stock closed at $51.32 on Friday, September 22 for a market cap of $52.7B. Intraday high during this period was $69. If SoftBank really did buy an additional stake in ARM at a $64B valuation, they’re down 18% with the P/S and P/E ratios still at nosebleed levels.

          1. Finster says:

            NVDA has been soft all this month, peaking at $493.55 August 31 and closing this week at $416.10. CART (Instacart) spiked to $41.08 on its debut Tuesday, closing this week at $30.00.

            Broadly it’s been a weak month for stocks; seems the speculative fires may be cooling.

  3. Finster says:

    Now it’s Instacart. Said to have gone public at a premium share price. I’ve used Instacart’s service, but in contrast to Nvidia and Arm, which are merely great companies selling at way too great prices, am skeptical of its business model. Like other services that rely on gig workers putting expensive miles on their cars, it’s not clear where sustainable profits come from. Eventually many workers, who initially see only their fuel costs, will factor in maintenance, depreciation, insurance, etcetera and discover they’re earning barely enough to cover their vehicle expenses, in effect jointing shareholders in subsidizing transient profits for insiders.

    So this is all about eager investors subsidizing businesses through the capital markets. Grab market share at a loss in order to crush the competition in the hope of recouping it later. Where is the intrinsic economic value addition? My small town grocer can deliver groceries just as efficiently as any big corporation.

    IPO waves tend to happen in markets frothy with eager buyers as sellers see they have the advantage. So I see this as a bearish indicator. Bloomberg reports the company founder “is checking out with a $1.3B fortune”.

    Does IPO mean Initial Public Offering? Or It’s Probably Overpriced? Or Insider Profit Opportunity?