Foreword

I know what you're thinking: another book?! What could Antti possibly have forgotten to say in his first 550-pager? I mean, you did read every page including the footnotes and the rather weak phoned-in Foreword, didn't you? Of course you did. Well, simply put, a lot happens in 10 years. Things happen in markets. Things happen in politics. Things happen in research. (Yes, we actually learn some stuff, and unlearn some stuff we thought we knew.)

For one, and you may have noticed this from the title of the book, markets have near ubiquitously gotten even more expensive. This has some potentially depressing consequences for the future, and, frankly, no option for dealing with this problem is particularly pleasant (save, we're all wrong, and stocks and bonds are actually dirt cheap right now - but don't hold your breath!). Globally, lower expected returns have no easy fix. But Antti, with some help from Stoics and St. Augustine, tackles this head-on in Part 1. Essentially the options are (a) take more risk so even with lower expected returns you can hit your goals (certainly counterintuitive as it's literally saying, "This looks worse than normal, so give me more," but some investors, with binding expected return floors they can't fall below, sometimes, hopefully reluctantly, need to do this); (b) ignore it and accept lower expected returns for the foreseeable future, don't "get cute" and try to ameliorate the problem, and just ride it out (I call this the "Jack Bogle" argument though, through our friendship, and with a twinkle in his eye, even Jack bragged to me a bit about selling some stocks in 1999-2000); or (c) find and incorporate other sources of expected return that either aren't very low now, or perhaps are compressed but are not correlated with the ones you invest in already (i.e build a better portfolio). Antti considers all these.

Another thing that happens after 10 years is - wait for it - we collectively have ten more years of data/history to learn from. While ten years is not enough time to seriously change our view of long-term expected returns (see my Foreword to Antti's first book on how hard this is), it's still the case that for some asset classes and strategies (e.g. illiquid ones), ten more years is a big fraction of their total histories, and thus it is still a pretty big deal to obtain them. And for other asset classes and strategies we have (not just Antti and me or AQR, though we've done our share, but the broader community of researchers) spent some of the past ten years building meaningfully longer histories by going back further in time. It's sometimes counterintuitive, and we recognize that older data may or may not be as relevant as current data, but as long as you've not yet peeked, new data further back in time is as much "out of sample" as future data. The credit premium, commodity premium, and especially style premia (aka alternative risk premia) are helpful data my colleagues and I (again it was not just us) are proud of having extended historically further back than previously known or at least widely available - in some cases now having nearly a hundred years of evidence (and in some really special cases even longer than that). To our satisfaction, and admittedly relief (you never know when going out of sample!), the older data provided yet more evidence that these premia are likely real and significant.

But, as excited as we are about more data - and you might not expect to hear this from a "quant" - more data doesn't always mean more "truth." Realized returns are noisy beasts over even time frames we'd all call long-term. As I've recently written about, changes in valuation can influence our estimates of realized (and expected future) returns a lot, over some surprisingly long time frames. (In one piece I show strategies like long-short value and just passive stock market exposure, starting out expensive and ending up even more expensive, have substantially distorted estimates of the natural expected return over even a seventy-year time horizon; see Asness (2021)). This Foreword is not the place for the details, but put simply, we believe that obtaining estimates of expected return that do not give undue credit for a strategy getting more expensive or undue blame for it cheapening (as rarely are either expected to occur in perpetuity going forward) is doable, yielding less biased (more accurate) and more precise (an underappreciated advantage) estimates for the future. But, none of this changes the sentiment that strategies getting very cheap or very expensive over the period studied do matter a ton in the real world over time horizons investors care about. We have often used the physics-envy term "time dilation" to refer to how long real life can seem while living it versus how short it can seem when checking out a backtest. It's easy to look at a good backtest in an intuitive robust strategy and examine its three- to five-year painful periods and think, "Of course I'd stick with it, it makes economic sense and look at the whole history!" But, I'm guessing, to none of our readers' surprise, it's a little harder to do live and real-time! To state the obvious, you need the best strategy you can stick with, not the even-better-in-theory strategy that you can't. You can influence that in two very different ways. You can alter your strategy, or come up with ways to put it in perspective that can help you stick with the best versions. We hope that, in particular, by explaining how distortive valuation changes can be, and coming up with concrete and useful ways to incorporate this into our estimates going forward, we can get both less biased and more precise estimates and, through the power of this argument, buck up some investors through deeper understanding. Luckily, Antti is a master teacher. In particular in Part 2 Antti brings his "twin perspectives" to bear, sharing tons of evidence, theory, and lived experience (we've both been doing this a while!) to help.

Antti and I have had many of the same teachers, from our days in graduate school to even our decades apart as researchers in the "real world," and especially learning from our colleagues over the past 10 years at AQR (not to mention AQR's "extended family" of academic consultants, co-authors, etc.). There's no getting away from the fact that this book is to some extent a reflection of Antti's beliefs, of my beliefs, and of AQR's beliefs. I say that for both disclosure, and because it's just true (not always the same thing!) and something readers should know. While Antti and I don't agree on everything (for instance, he's Finnish and thinks going from a 190° Fahrenheit sauna to immediately roll around in the snow is "healthy"), we do agree on much more than we don't.

And this brings me to the uncomfortable (for me) topic of Antti sharing our stuff. Antti is, to put it politely, an "over-sharer." This is generally a good thing in a researcher. No hiding assumptions, lots of musing about the many reasons he could be wrong with an open mind, all good things. But, the crux here, he has a true natural bias to reveal some things we'd sometimes prefer to keep to ourselves for a while! But that's my problem, not yours. I think it is certainly good for you as the reader! Though, to be fair, we don't let Antti share everything. He goes pretty far herein, even further than his last book, but we do believe we have some sources of "alpha" (such a loaded word) that are still relatively unique and there we do clamp down on Antti a bit (not an easy task!). Maybe, if the pattern continues, we'll let him go even further for book number three in the year 2031.

OK, back to the book. Even though it starts with doom-and-gloom in terms of low expected returns, it doesn't end that way. Parts 2 and 3 cover a wide range of ideas to improve investor outcomes that don't rely on markets going up. Some are strategies not particularly rich or cheap (so not a tactical view) but just sources of expected return not correlated to the major ones and, we believe, endemically underutilized in most portfolios. But some are tactical. In particular, what we have deemed (back in 1999-2000 for you old timers) the "value spread" shows that, unlike so many other premia that appear expensive, the value premium, historically positive on average, seems record cheap now (a self-serving opinion and one that will date my Foreword!). Antti covers this (e.g. in Figure 6.3 and a few other places). But, given it's one of the only things I can think of that both has made money over the very long-term and is currently very cheap versus history, it deserves its own mention. When such a reversal will happen is, of course, always the most difficult part. But that we believe it will happen net from here is one of the only silver linings we find among a world of almost all low expected future returns, and, again, although self-serving (I got my kids' money way overweight on this one and they're starting to ask for portfolio reviews over dinner!), I think it's exciting enough to stress here even more.

Finally, there are also some often overlooked parts of the investment process - not as sexy as what to buy and when to buy (or sell) it, but still vital: things like risk management, portfolio construction, and even the mundane but vital area of trading costs and fees. Antti is steeped in both real world...