MSFT - Value's U-Turn

2023-11-10 07:30:00 ET

Summary

• Value went through the longest and deepest dry spell in history from 2007 to summer of 2020. Depends how you define value.
• As a value investor, obviously that puts our company through a ringer and it has us challenging and testing our own assumptions.
• Growth stocks are more expensive. They're better companies and they deserve a premium multiple.

Rob Arnott, Founder and Chairman, Research Affiliates

Interviewed by Jamie MacDonald, ex-portfolio manager, Point 72

Jamie: Rob, I want to talk about Research Affiliates and the products you have on offer. But before we do that, we've spoken about growth versus value before. And me included, there were many of us who felt that growth had had its time in the sun and with higher rates, this was going to be a period for value to outperform. We have got the higher rates. It appears they are going to be here for a while, but still, there's a lot of tailwinds in growth.

Rob: A very narrow sliver of growth. Big players, but small number of names. So actually, what we're doing there is, we're talking about growth as a whole, but it's just a handful of the Nvidias of this world that are just taking that whole segment of the market. I read the other day that the five most successful stocks this year in the S&P are the only reason it's up year to date.

Jamie: Is that right? I haven't seen that maths before.

Rob: Yes. So value went through the longest and deepest dry spell in history from 2007 to summer of 2020. Depends how you define value. If you define it on price to book, it was that full 13 years. If you use price earnings, it was from 2013 to '20.

If you use price to sales, 2017 to '20. Either way, it's a long dry spell and the 2018 to '20 period can only be described as a value crash. Value underperformed severely.

Now, as a value investor, obviously that puts our company through a ringer and it has us challenging and testing our own assumptions. So we wrote a paper back at the beginning of 2021 in the FAJ. It got Graham and Dodd recognition as one of the two best articles of the year.

Jamie: Oh wow.

Rob: The paper was entitled, "Reports of value's death may have been greatly exaggerated." The basis of the paper was research project asking, has value in fact died? Has it died? Are the critics right? And what we looked at, among other things, was the spread in valuation between growth stocks and value stocks.

Now if you use a classic Fama-French formulation, you're using price to book. And in 2007, the spread in price to book between the growth stocks and the value stocks using Russell Growth and Russell Value for instance, was four to one.

Growth stocks are more expensive. They're better companies and they deserve a premium multiple. But how much of a premium is the key issue. That went to 13 to one by the summer of 2020. It only reached 10 to one at the peak of the tech bubble. So this was tech bubble on steroids magnified.

Now, what's interesting is that Russell Value peak to trough underperformed Russell Growth in terms of relative performance by 3,700 basis points over that 13-year span.

Russell Value got cheaper relative to Russell Growth by three to one. So if it's down 67% in relative cheapness and down 37% in relative performance, then that means the underlying fundamentals have actually been improving relative to growth. I'm going to show an exhibit that shows this vividly. The dividend stream of a Russell Value portfolio from 2007 to 2022 has risen faster than the dividend stream of a Russell Growth portfolio.

Jamie: Really?

Rob: And it's not because the value stocks grow faster. It's because of what Fama and French called the migration effect. Value stocks, the value portfolio, you have individual stocks that percolate out of it suddenly get recognised as, "Oh, this company's not so bad after all."

They're kicked out and replaced with deep value stocks. So if you replace a stock with an earnings yield of five, P/E ratio of 20 with one that's earnings yield of 10, P/E ratio of 10, then you're boosting the earnings base, you're boosting the dividends base, the book value base every time you rebalance. It's called a migration effect because it relates to migration in and out of the list. The growth portfolio has the opposite.

Stocks that are removed from growth almost always because they've fallen to too cheap evaluation. So, if you replace a stock with a P/E ratio of let's say 15 with one with P/E ratio of 30, you're reducing the income stream, the dividends, the book value of the growth portfolio.

So, net of the migration effect or inclusive of the migration effect, I should say, the dividend stream for value, the book value for their value portfolio all grew faster than for the growth portfolio during the period of time when the value was getting crushed.

Which means if the relative P/E ratio, relative price-to-book, had stayed steady for those 13 years, value would have beaten growth. So, we pointed this out, we also pointed out the deep flaws of using conventional book value. And the result was a pretty compelling case that value stocks were doing badly, but value companies were doing fine.

Rob: Now, markets move because of narratives. The narrative was with the COVID lockdowns, these tech companies are beautifully positioned for a world in which people work from home, in which people don't interact as readily as they used to.

A world in which social interactions are going to change, the way you buy and sell goods is going to change. And the list goes on. And by the way, these bricks-and-mortar companies, they're toast.

You're going to see rolling bankruptcies. Well, lo and behold, with stimulus checks paired with a non-permanent pandemic lockdown, we found that the value stock survived just fine.

Bankruptcies in 2020 were up modestly from 2019. Shocking that it was only a modest jump. And so, the result was that although the market was saying these value companies are going to do terribly as businesses, the reality was they were doing fine.

The snapback is something that we were predicting in the paper. What we weren't predicting was the current snapback in growth year to date this year. And that's on the back of AI. The narrative today looks an awful lot to me like the narrative of the year 2000.

Jamie: Right.

Rob: Back then you had internet about to change the world, a new paradigm. Pay no attention to profits or dividends, because they're irrelevant. These companies are going to take over the way we do everything. Now the narrative is that AI will do that.

Jamie: Do you feel we're in a bubble then?

Rob: Yes. But bubbles exist because narratives have the advantage that they're almost always at least partly true and often very true. But they have the drawback that they're already in the price.

Jamie: Right

Rob: The current price reflects the accepted narrative. The accepted narrative right now is that AI is going to change the world in remarkable ways in the years ahead. I agree. If you've played around with ChatGPT or Dolly or any of these tools, it's astonishing what they can do.

But the same could have been said about the internet in 2000. And I just completed a three-week trip in Europe. Is AI going to change the way I'm driven around in cities? Yes, it’ll take time, but yes. The baggage handlers, not really. The chefs, the waiters in restaurants, the people who show you up to your room at a hotel, I'd say 80% of the people I interacted with, I was watching and thinking, "Is this person's job at risk on a five-year horizon?" 80% of them, I'd say no. Which basically means this is going to be a monumental revolution gradually. More gradually than people think. If I was a web designer, I'd be in a panic.

Jamie: Yes, or a coder.

Rob: If I was a mediocre coder, I'd be in a panic. If people talk about three to 400 million jobs disappearing in the years ahead, pardon me, but that's actually fairly normal. The US loses over 2 million jobs disappearing every month and two and a quarter million new jobs created.

Jamie: That's just the natural course of evolution.

Rob: That's the natural evolution of business. And so as with horse-drawn carriages being replaced by automobiles, lots of jobs. The cliches, buggy whip manufacturers. But the list goes on and on. And is that horribly rough on the individuals whose jobs are affected? Of course, it is.

We should, as a society, be humane about that. But it doesn't mean saying you never have to work again, we'll take care of you. It just means we'll help you help yourself. So the AI revolution is very real. Does that make Microsoft ( MSFT ), with its close tie with OpenAI, does that make them worth 12 times sales? I don't think so. That's a huge multiple of sales.

Jamie: Yes. So difficult to know what the sales number would be. You're just printing multiples on things you couldn't even possibly predict.

Rob: Well, two jumbo-cap companies, Microsoft and Exxon Mobil ( XOM ). Exxon Mobil, best of my recollection, is less than two times sales. So tacitly, that means that people expect that Microsoft's gross revenues are going to grow six-fold relative to Exxon Mobil. That’s big.

Jamie: So, Rob, if you could explain to me the partnership products you have with FTSE Russell and RAFI.

Rob: FTSE embraced the fundamental index concept way back in 2005. We brought the idea to FTSE at the suggestion of CalPERS, who embraced the idea. FTSE was the index provider for CalPERS. And so obviously, we had high odds of a near-term immediate client.

So FTSE embraced the idea, ran with it, and it became a major profit engine. Russell came to us and said, "We're going to launch similar product. Do you want to work with us on it?" And my original agreement with FTSE had, given a three-year exclusive, was six years in.

So I said, "Sure, but it has to be different." They initially wanted it to be a clone and I said, "No way." So we created a complementary, somewhat different, but think of it as first cousins. FTSE, of course, bought Russell. And so now, they're all under one roof. So you have FTSE RAFI and Russell RAFI, and upwards of a hundred billion managed using them.

Jamie: I'm right in saying performance has been pretty great since COVID.

Rob: Not since COVID. Since the launch of RAFI. To the casual observer, RAFI had a challenging decade in the 2010s. But keep in mind, cap-weighted indices studiously mirror the look and composition and performance of the stock market. By definition.

Jamie: That kind of momentum plays.

Rob: But relative to the economy, their momentum plays, their growth plays their popularity-weighted indices. RAFI is studiously neutral relative to the publicly-traded macro economy. It weights companies according to how big is their current economic footprint right now. And so, relative to the economy, cap-weighted indices are stark growth portfolios.

Relative to the market, RAFI is a stark value portfolio. So, you can do a Fama French attribution and look at the alpha net of Fama French, or you could play it super simply and just say, "Okay, the value tilt of RAFI roughly equals that of the conventional value indices on average over time." It's dynamic.

When value underperforms, we have a deeper value tilt than value. When value is relatively fully priced as it was in 2007, we have a pretty skinny value tilt. But on average, about the same value tilt as value indices. So, the tracking error of RAFI Global against FTSE All-World is 5%.

Against All-World value, it's 2.5%. So, if you look at it relative to the value indices, what we find is since the launch of RAFI during periods when value was winning, during periods when value was losing, during the value crash as it was happening, we beat the value indices by an average of 1.5-2% per annum. Now for 18 years, with 2.5% volatility.

Jamie: That's great.

Rob: It's already got a t-statistic of three-plus. So these indices have been stupendously effective. The challenge in the last decade has been a marketing challenge. People will inherently compare it with the cap-weighted markets.

And whenever value's losing, we may be losing by less, but we're still losing. When value is winning, we're winning by more. So we're perceived as heroically hitting the cover off the ball. Now, how long can that continue? When if you compare it to the value indices, you find it's just chug, chug, chug, chug relentless.

Jamie: Right. I have one final question. You said earlier, Rob, that you wrote a paper talking about the valuation metric price to book, how you didn't think it was such a great metric to look at. Am I right?

Rob: It's flawed. It's not a bad metric, but book value was defined either early 20th or late 19th century.

Jamie: The reason I'm asking is I'm wondering if now, is there a better way to value stocks?

Rob: Oh absolutely. We use a composite of measures. If you look at price to sales, price to book, price to cash flow, price to dividends plus buybacks, these are all good measures, and they are all flawed. So George Box of Box-Jenkins fame was fond of remarking that all models are wrong, some models are useful. And oh, I wish the quant community and the economics profession would embrace that.

Everything we do in the quant community, everything that's done in the economics community is based on models that are wrong. But some of them are useful. So find out which ones are useful, use those. The pitfall with book value is that it leaves out all intangibles. We've all heard the cliche that our assets go up and down in the elevator every day. That is true of about half of all businesses today.

Rob: The old bricks and mortars, businesses, a department store, it's not as true. The assets are the people for sure, but also the building and the product that they're selling. And that's a big sunk cost. So book value measures that. Now, what if you take book value and when a company invests in R&D, you add it to the book value and amortize it out.

If I spend five grand on a nice desk, it goes right on the balance sheet as an asset. It's amortized out. If I spend 5 million on R&D, it's treated in accounting land as if it was just a discretionary throwing away or burning of assets. Now, I don't spend that on R&D if I think it's going to not come back. And so think of it like a desk, add it to the book value, amortize it out.

If after 10 years it hasn't paid for itself with room to spare, then it was a dumb idea. Same as a desk that turned out not to have needed. So when you do that, Fama French price to book, the value investor today is five times as wealthy as the growth investor from 60 years ago.

So, if your grandparents bought Fama French value, their next-door neighbour bought Fama French growth, you would've inherited a portfolio that's worth five times as much as the descendants of the next-door neighbour. All right, well, that's great. If you threw in just R&D, nothing else, just R&D, you would have 10 times as much money.

Rob: So, price to book can be made roughly twice as effective for the long-term patient investor by incorporating R&D. You can also incorporate a portion of the administrative side to the extent that somebody spends money on patents, on brand building through advertising, and so forth.

Those expenditures are justifiably treated as investments and therefore added. But that's trickier because that's all-in baskets that also include ordinary operating expense. And so, the one piece that is easy to carve out and say, "Well, this is investment," is R&D. So that was a, in that same paper, it really could have been two papers.

Jamie: Well, such an important thing to discuss right now as people try to work out whether stocks are overvalued or not. Rob, it's been so fantastic chatting to you. Thank you so much for your time. It's always such a great conversation.

Rob: Well, I have thoroughly enjoyed this nearly two-decade long partnership, and I look forward to many years to come.

Original Source: Author

Editor's Note: The summary bullets for this article were chosen by Seeking Alpha editors.

For further details see: