Wednesday 11 November 2020

Unravelling value's decade-long underperformance (and imminent resurgence)

In a recent (generally excellent) podcast with Inside the Rope with David Clark (#78), John Hempton discussed (amongst other things) value's past decade of underperformance, and opined that the primary driver was the fact that the pace of technological change had accelerated, such that we have seen an unprecedented level of disruption to traditional business models. Value investors have apparently spent a decade naively riding doomed low-multiple companies like the Myers of this world into oblivion. 

This is a very commonly expressed view/belief, and intuitively it feels right. The danger with intuitively-satisfying beliefs though is that they can discourage you from looking for evidence to confirm whether those intuitions are in fact true. It seems true, so it must be true, right? A surprising amount of the time, the answer is no. Just because something is intuitive does not mean it is correct (after all, it was intuitive to pre-modern humans the world was flat, and it's not very intuitive we evolved from primordial sea creatures), and as Mark Twain once noted, "It ain't what you don't know that gets you into trouble, it's what you know that just ain't so". The story of the emergence of the scientific method is a story of humans starting to demand evidence instead of merely relying on our unsubstantiated intuitions.

Cliff Asness of AQR has undertaken a comprehensive systematic quantitative analysis of this argument to determine whether this intuition is correct, and found absolutely no evidence to support its veracity. Contrary to popular belief, there has been no degradation in returns on capital or earnings for value quintiles, which would substantiate the existence of excess 'disruption' in value as compared to historical averages. In fact, value portions of the market have actually done slightly better on these metrics vs. long term averages over the past decade. Asness' analysis concludes that the primary driver of value's underperformance has simply been value getting cheaper and growth getting more expensive, as has been the case in every past cycle where value has underperformed (of which there have been many).

It is very easy to forget that disruption is not a new thing. It's been going on for as long as the forces of creative destruction have operated in capitalist economies, which is to say for at least as long as stock markets have existed. Technology and the competitive landscape has always been evolving, and upending certain business models. The automobile industry disrupted the horse and proverbial buggy whip industries. Electricity and the incandescent lightbulb disrupted kerosene lamps and candles. Growth of low cost manufacturing in Japan, Korea and other emerging markets disrupted many traditional manufacturing industries in the US in the 1970-80s, from appliance and automobile manufacturers, to steel producers and textile millers. Retail has been through many eras of disruption. such as the emergence of the category killing Wal-Mart big box discounters of the world, who disrupted the dominance of traditional department stores (who had previously disrupted smaller independents). 

The only constant over time has been continuous change, and value has always faced the headwind of certain companies/industries being disrupted - often fatally. It's not a question of 'do some cheap stocks end up getting cheaper - perhaps all the way to zero - and justify their cheapness', but rather, to what extent does that happen in the aggregate, and to what extend does the lower entry multiples for these stocks compensate for that risk, including not only the benefit of higher dividend yields/more rapid cash flow paybacks, but also the benefit gained from exposure to situations where disruption fears prove excessive and stocks recover. Hempton's argument is that the pace of change/disruption has accelerated, but Asness' quantitative analysis does not support that assertion - at least from the perspective of it having a tangible impact on value stocks' aggregate fundamental performance. 

Furthermore, it is a major mistake to assume the impact of disruption is confined merely to low multiple stocks (or even felt disproportionately by value). Kodak was a very highly rated, high quality company up until the late 1990s, as was Blockbuster video rentals. That didn't stop them from being disrupted. Indeed, it is actually often the highest quality and highest rated companies that have the most to lose from disruption, as they have both high valuations with very long duration payoffs and very high profitability. This means not only do they have a long way to fall if anything goes wrong (and even the fear of disruption can crush these stocks, whether or not it actually transpires), but their fat margins also act to invite disruption by creating an outsized opportunity for would-be disruptors. One of the reasons Uber exists is that taxis were previously morbidly overpriced, and one of the reasons we have not seen (and are unlikely to see in my view) fintech disruption of the banking industry is that lending spreads are already very thin, and the industry highly capital intensive (onerous regulatory capital requirements) and not especially profitable, so there is little opportunity/reward for doing so.

Examples abound of formerly highly-rated franchises that were part of the high multiple growth and quality parts of the investment universe that suffered massive disruption and falls from grace, and the tendency of a not immaterial minority of highly-rated quality/growth stocks to fail to live up to their expected potential and suffer major de-ratings has been a fundamental driver of the long term underperformance of high multiple stocks as a group. The idea that it is the value portion of the market that is most exposed to disruption is therefore a case that is far from made - intuitive though it may be. By the time fallen angles end up in value bucket, it is right to say that the probability of catastrophic disruption has materially increased, but on the flip side, by this stage the stocks have already handed major losses to their prior growth/quality holders, and have now already de-rated to the point where investors have very low expectations of the company's future, and offer high dividend yields/more rapid cash flow paybacks as risk mitigations. And if the market is wrong about the inevitable future demise of such companies, longs stand to make many multiples of their investment.

I have argued many times on this blog that one of the core reasons that value investing works is that investors systematically overestimate their ability to predict the future. Now sure, we have the J.C. Penny's and Sears of this world that slide slowly but surely into bankruptcy, as their low multiples long predicted would happen. However, you also have examples such as electrical appliance retailer Best Buy that was predicted to follow a similar trajectory, but which instead managed to successfully reinvent itself, leading to its stock rising 10x as its future prospects were reassessed. More recently, stocks like Fedex have risen 150% from lows of below 10x earnings as fears of Amazon disrupting their network advantage have abated.

A decade ago, investors believed Microsoft was being disrupted by mobile, Apple & iOS, Google Docs and Android (Hempton even blogged about how he went short at the time after watching a Youtube video where someone's dad was totally baffled about how to use their latest version of Windows). The stock traded for as little as 7x earnings at the time. Suffice to say investors' predictions of Microsoft's imminent demise (a value stock at the time, I might add - something today many people forget) were more than a tad exaggerated - as well reasoned as that case seemed at the time. The same can be said of Apple, which was losing money and widely believed to be heading for the bankruptcy courts in circa 2000 - again quite reasonably based on what was knowable at the time - and the stock accordingly traded for less than its net cash. We know how that ended. Yet another example is the very industry Hempton discusses he is buying at present on the podcast - the tobacco industry. Long thought to be a dying industry and priced as such, it went on to be one of the best performing industries in the world over the past 30 years (I agree with Hempton here and am also long BAT, and blogged about it here).

So sure - there are individual examples of low multiple stocks being disrupted all the way to zero, as there always have been. As Buffett will tell you, the US textile industry got disrupted all the way to zero during 1970-80s from cheap imports. But there are also plenty of examples of stocks going up 10x or more as those disruption fears failed to materialize as expected, and with this degree of asymmetry the probability of the market being wrong doesn't need to be very high to generate a favourable return expectancy. A century of quantitative evidence from market history suggests investors tend to underprice stocks with the most apparently assuredly poor future prospects, and over price those believed to have the most assuredly promising prospects, and underestimate tail risk (both upside and downside), and there is nothing in the past decade's market experience to suggest that has fundamentally changed. 

Further evidence of this stems from the multiple studies that have been done on net-nets - the worst of the worst in terms of business quality and future outlooks (the outlook is so assuredly bad investors are not even willing to pay a price above net working capital net of all liabilities). As a group, such stocks have substantially outperformed over time. However, very interestingly, when studied have been done where investors were given the opportunity to choose the 'best of a bad bunch', choosing only those that were profitable or paid a dividend for instance, the results were much worse. Taking out the 'worst' of the worst lead to inferior returns. Why? Because if it's obviously bad to you, then it's obviously bad to everyone else as well, and the stock will be priced accordingly, with the probability of unexpectedly favourable change underestimated, leading to greater scope for a major re-appraisal of its prospects if conditions do unexpectedly improve. And occasionally, that happens. Most of the time it doesn't, but sometimes it does, and occasionally you end up with an Apple (which was a net net circa 2000).

One of the difficulties investors struggle with is the asymmetry. I've blogged about several value stocks on this blog before. Lot's of them haven't worked out particularly well, although in most cases they are flattish or only modestly down (inclusive of dividends). However, some of them like Fortescue are up 500% including dividends. The issue with value stocks is the high dispersion of outcomes - lots of stocks deliver mediocre outcomes, and handful go down a lot, and then some deliver very extreme payoffs. This makes it difficult for investors to learn the right lessons as they will often have bad outcomes on individual positions even if the overall ex ante return expectancy was high, and critics of the approach will always be able to point to individual situations which didn't work out in an attempt to discredit the methodology. Pointing to specific value names that got disrupted all the way down to zero as a reason why value investing is flawed and doesn't work is doing precisely that.

So if accelerating disruption is not the cause of value's underperformance over the past decade, then what exactly is? The simple answer is because cheap stocks have got cheaper while expensive stocks have got more expensive - a trend which has radically accelerated over the past 24 months or so (barring the sharp reversal we have seen over the past few days). However, the more complex and interesting question is why and how that has happened, the answer of which requires that we go all the way back to value's last period of significant underperformance in the late 1990s, as well as understand the important role liquidity flywheels (discussed in detail here, and a recommended pre-requisite to the below analysis) and investor psychology play in long term secular market cycles.

It is worth emphasizing at the outset, however, that the tendency for value to lag markets for significant periods of time is not itself historically unusual, and happens whenever a market environment exits where expensive stocks get more expensive; cheap stocks get cheaper; or some combination of both (as we have seen this cycle). Over the past 50 years, it has happened three times in US markets - in the nifty-fifty bubble of the 1960s-early 1970s, in the 1990s bubble, and over the past decade. In other words, it has happened approximately once every 20 years or so. The only thing unusual about this cycle has been its length and magnitude, which has been somewhat worse than past cycles.

The fundamental causes have been the same. You start with a market environment where value has done relatively well, and multiple dispersion has reduced. If multiple dispersion is low, it actually pays to own higher quality, higher growth companies, as you're not paying much extra for these benefits. Subsequently, there is some sort of new trend/theme that emerges which leads to a cohort of companies performing unexpectedly well operationally (robust profitability and rapid growth), and their stocks perform exceptionally well, as they benefit from a combination of a reasonable starting multiples, operational prosperity/growth, and then a significant lift in trading multiples as their success becomes more widely appreciated as it is promoted by brokers and the media, and investors start to extrapolate and price in continuing growth and prosperity forward into the distant future. 

In the nifty-50 bubble it was high quality US blue chips that benefitted from a strong economy in the 1960s and low interest rates, as bonds lost their appeal and investors substituted high quality blue chip stocks which had reliable earnings power and steadily growing dividends. The stocks did so well for so long as multiples rose and rose from 10-20x to eventually as high as 60-80x that they became known as 'one decision stocks'. In the 10-15 years that followed, they were a total disaster as multiples reset back to 10-20x and the stocks fell 80%. In the 1990s it was the emergence of personal computing, software (enterprise and personal) and the internet, which produced some spectacular secular winners such as Microsoft, Cisco, Oracle, Intel, and Dell. Investors made a fortune in tech during the 1990s, until the tech wreck where the NASDAQ fell 85%; and post-GFC, it has been smartphones, social media, cloud computing and enterprise SaaS, e-commerce and various online consumer marketplaces, which have produced this cycle's cohort of secular winners such as the FAANGs and SaaS leaders such as CRM. 

What all of these cycles have in common is that the initial bout of outperformance was fundamentally justified by emerging secular trends and reasonable starting-point valuations, but subsequently, as a liquidity flywheel was set in motion that drove rapid multiple expansion over many years, the trend ended up being carried to morbid excess. What happens is that fund managers that due to good luck or good foresight owned those secular winners early report great numbers, and great numbers attract inflows. Those inflows are then invested in the same names, pushing share prices higher still. 

Investors' greed and get-rich-quick instincts are piqued by strong and consistent performance, and particularly when buttressed by an exciting thematic narrative that seems to justify the strong gains and promise more to come, and with results appearing to validate that assessment. More sector-based funds are birthed and promoted to cash in on this growing investor enthusiasm, and as more and more money flows in, prices get pushed ever higher, further validating the narrative, emboldening investors, and dulling risk aversion. And because investor greed is insatiable and accelerates alongside prices, this self-perpetuating cycle can go on for a long time - many years - fermenting a bubble. As is often the case, Buffett said it best when he said "What the wise do in the beginning, fools do in the end".

It is sometimes said that the definition of a bull market is a random market movement that causes investors to mistake themselves for financial geniuses (ARK's Cathie Wood is a great example of an investor suffering from that affliction at present). Powerful and prolonged liquidity flywheels cause many investors to mistake their significant gains for investment prowess and accurate foresight of secular trends, when in fact they are being driven by the forces of liquidity. I can articulate every single secular disruptive trend Cathie Wood espouses just as persuasively as she can. But I'm not under any illusion that these fundamentals are what is driving current share price gains. Wood would have been loaded up on AOL in 1999 talking about how the internet was going to change the world.

Anyone that has invested in a successful growth company during a secular bear market will understand this - such companies can deliver outstanding results year after year after year and the stocks can still go sideways/down. There are market-leading companies I know in Africa that have grown at 15% a year for the past 5-10 years, with high returns on capital and growing dividends, and yet seen their their stocks fall cumulatively by 50% (from about 10-15x earnings to 2.5x earnings) simply because frontier markets have suffered outflows, and accordingly there are more sellers than buyers for the stock, and price declines trigger yet more sellers (redemptions), so stabilisation/equilibrium proves elusive. For often surprisingly long periods of time, it is liquidity which drives market prices, not fundamentals. Anyone that owned Amazon during 2000-03 where the stock fell 93% will also know this. 

As a liquidity driven boom roles on year after year, investors become increasingly skeptical about the role of valuation, for the simple reason that valuation has proven to be a poor predictor of share prices in recent history. Stocks that looked expensive just kept going up (due to liquidity, which is why they were expensive in the first place), so investors - many of which lack decades of experience - come to believe that focusing too much on valuation is a bad idea. Investors will also point to a handful of big secular winners like CSCO and MSFT (in the 1990s) and AMZN this cycle and note they were 'always expensive' and that it was a mistake to pass them up simply because they didn't trade on low multiples. They will then use this logic to justify paying almost any price for companies of vastly inferior quality, ignoring how unique and uncommon companies like AMZN are, so long as stock prices keep going up and validate the narrative. They are right that valuation is not a good predictor of share prices, but are wrong about why. They think it is because it is growth and business quality driving returns, when in fact it is simply liquidity. Nifty-50 investors learned this the hard way when the same high quality businesses with the same high quality and defensive operating results they had always had fell 80% in the 1970s.

As the boom roles on and the market capitalisations of favoured names mushroom, people lose all sense of scale, or at least choose to ignore it. In less heady times, investors worry that very large capitalisation stocks have less room to grow and so hence offer comparatively limited opportunities for large share price gains. Late in a liquidity driven boom, large market capitalisations are seen as no obstacle to continuing sustained outsized returns, because such companies are capable of driving rapid growth 'at scale'. The problem is that the world is large but it is not infinitely large, and trees don't grow to the sky. 

For instance, Zoom Communication's peak market capitalisation was recently about US$200bn. Even to trade on a relatively high 20x earnings, it would need to earn US$10bn after tax. Are investors aware of how few companies there are in the world that actually make US$10bn? It's about as much money as Coca-Cola and Visa make, for instance - two of the world's finest enterprises. Very few companies make more than US$10bn, because that is a lot of money, and the world is not infinitely big. And yet I'm convinced most of the people buying Zoom at those prices actually expected to make very large and rapid returns, because, ya know, the company is growing really fast.

This sort of foolishness was widespread in the bubble. Near its peak, one analyst pointed out that AOL's market capitalisation already priced in more than everyone in the entire world having an internet connection. Investors weren't paying attention to the inevitable limits to growth because growth was really fast right now, and looked likely to remain fast in the foreseeable future, and well, ya know, the internet was going to be huge. This was able to happen because investors were so preoccupied with how much money they were making and how big of a growth opportunity the internet was going to be that they stopped doing any valuation math. Why? Because any argument that a stock was too expensive led to the 'wrong' investment decision being made over the past 5yrs+ as prices kept going up.

As the boom roles on and liquidity mushrooms, a wave of second-tier IPOs emerge, greeted by rampant investor enthusiasm, as investors seek to catch the 'next' secular winner (like the 'next Microsoft' in 1990s, or the 'next Amazon' in our times). Investors will justify paying virtually any price for companies like (for e.g. Sea Ltd), ignoring competition and operating losses, 'because Amazon lost money in the early years too'. At this point, it doesn't matter that the huge gains these stocks are enjoying represent pure speculation and hope about future prosperity, despite the comparative lack of it today or any realistic assurances it will actually materialise in the future. Risk aversion dissipates because such a long period of rising prices and rapidly growing top lines make the risk of loss seem remote.

Peter Lynch observed that it's always incredibly dangerous in markets when investors say company Y will be 'the next X'. In his experience, Y almost always blew up, and in my view that is because outsized success requires a unique and unlikely alignment of stars that occurs infrequently, and is also often the result of a lack of competition leading to an early advantage. The bust for instance may have helped Amazon a lot by cutting off access to capital to new emergent competition for many years, giving it time to solidify its lead. That doesn't happen in an environment where a million startups are getting funded and VCs are throwing billions of dollars at anything with a large TAM. When you have half a dozon companies all throwing billions of dollars at becoming the 'Amazon of South East Asia', a far more likely outcome is that they all fail and simply end up incinerating cash battling it out amongst each other for market share, just like the ill-disciplined airline industry of old.

At the late/extreme stages of a cycle, it can often reach the point where investors liquidate other assets wholesale in order to increase participation in the boom. This is usually the point in the cycle where multiple dispersion really starts to accelerate, and value funds not only lag from a relative perspective, but also begin to report poor absolute returns as well, as redemptions force sales and drive down prices.

In the late 1990s, the exact same argument was made about accelerating disruption being the justified cause of value's underperformance. Value funds' multi-year lagging returns led to them being discredited and suffering redemptions, as investors sought better returns elsewhere from managers that 'got' the new world order. Value funds were being shuttered left right and center, and Warren Buffett had one of his worst relative years ever in 1999, underperfoming by about 20%, to much mockery and condemnation that Buffett had become too old and was out of touch with the brave new world. Computers, software, and the internet was going to change everything. Value investors just didn't 'get it'. The truth, however, was that the starry-eyed growth crowd just didn't get it, and performance differentials were simply due to a rampant liquidity flywheel driving multiple dispersion to extreme levels. 

A lot of investors today like to argue that the current tech cycle is different because the FAANGs today are powerful and monstrously profitable businesses, whereas in 1999 we had and Webvan. But this is a major false equivalence. Today's FAANGs should be compared to 1990s secular winners such as Cisco, Microsoft, Oracle, Dell, and Intel etc, which were also monstrously profitable, and compounded earnings at 30-40% pa over a decade, while multiples levitated to 80x. These were globally dominant companies delivering rapid growth 'at scale' (as people like to celebrate today with FANGs), and were poised to benefit from (and dominate) the coming revolution in computing, software, and the internet. Fortunes were made in these names, and that is why there was an appetite for low quality IPOs - everyone wanted to buy into the 'next' Microsoft. or Webvan should be compared to low quality IPOs like Nikola today, which people are buying because they want to buy into the 'next Tesla' - not to Google.

Value funds were left for dead during the late 1990s not because value stocks performed particularly poorly or were disrupted, but simply because they didn't own these high fliers. This mirrors the performance of value this cycle up until about 18-24 months ago, where the results in absolute terms hadn't been that bad - they merely lacked the spectacular gains of some of this cycles secular winners. However, a tipping point typical of late cycle excess began to emerge 18-24 months ago where we began to see outright liquidations of anything non-tech in order to get more tech exposure, compounded by the coronavirus, which has only hastened the rush into secular tech winners and away from everything else. This has lead to rapidly falling multiples for value and exponentially higher multiples for tech.

The other thing the late stages of this cycle have in common with the era is reckless excess in the conduct of monetary policy. In 1998, the Fed overreacted to the Asian Financial Crisis and slashed interest rates at a time a mature tech boom was already well under way, which helped propelled tech stocks to their final ludicrous heights. We have seen much the same thing happen as a consequence of extraordinary monetary stimulus/QE during the covornavirus downturn. 

Of course, we know how the last cycle ended. Just as value investing was being declared well and truly dead, the NASDAQ proceeded to crash 85% and value investing enjoyed a resurgence. The bust happened for predictable reasons. The first thing was that investors forgot what happens every cycle - a boom leads to a massive flood of capital into an industry, and when this happens, supply and competition increases, which eventually drives down growth and profit margins. Investors believed there would be so much growth in software, compute, and telecoms/the internet that no amount of capital would be too much to overwhelm demand, and that industry players would be immune from the prosaic forces of competition and cyclicality that ailed the 'old economy'. They were wrong. Way too much fiber optic cable was built, for instance - so much that it took a couple of decades to absorb, despite the extraordinary growth in internet traffic we have witnessed. Instead of mushrooming growth and rising margins, we instead saw slowing revenue growth and a collapse in margins, as supply overtook even the torrid pace of demand growth, and competition heated up. This then led to a need to reign in costs and slow capex, as happens in every cyclical downturn, which led to cascading revenue pressures throughout the technology supply chain.

An analogous situation today would be Nvidea. Nvidea is a great company, with innovative products. However, it trades at 80x cyclically-high earnings, because people believe there will be so much growth in AI and other high-performance computing workloads that rapid growth and high margins will be forever unstoppable (as they thought with Intel in compute, or Cisco and networking kit, in 1999). However, as great as Nvidea is, it's growth is leveraged to the second-derivative of HPC/AI workloads - i.e. the rate of growth in the rate of growth. If there is merely a slowdown in the rate of growth in AI compute (i.e. it's still growing, just at a slower rate), Nvidea's revenues and earnings will fall. In a serious tech cyclical downturn, their revenues could halve and profits drop 75%. If the stock fell 90% it would still be on 32x P/E in this environment. That's pretty much what happened to Cisco and Intel.

The other thing that happened was that the stock market worked in fulfilling its capital intermediation role. If there is insatiable appetite for anything tech which drives valuations higher and higher, the financial industry will manufacture more product to sate that demand, which included a flood of tech IPOs. Eventually there was so much new IPO product it was able to absorb and overwhelm the wave of buying liquidity. We are seeing the same thing today. In the past on this blog, I talked about how there was a VC bubble unmatched by the stock market, and this was why we were seeing so few tech IPOs - the valuations would not stand up to the scrutiny of public markets. That has now changed - the IPO/listed space has become as/more frenzied than the VC space, and this has led to a flood of tech IPOs. As more and more IPOs come to market, not only is more capital raised to fund yet more product development and hence more competition, but there is simply are greater supply of stock to sate speculative demand. Secondary issuances, and continuing copious SBC (stock based compensation) and insider selling serve to further continuously increase supply. At some point, the force of supply will start to overwhelm demand, and that happened in 2000. And it led reflexively to an escalating cyclical downturn as tighter access to funding slowed IT spend, which had cascading impacts through the supply chain.

As the tech bubble burst, suddenly formerly-discredited value investors didn't look so silly anymore, and good quality companies making stable earnings and paying solid dividends also looked more appealing now that tech stocks weren't going up hundreds of percent a year, but instead handing investors losses of 80-90%, all while paying nothing in the way of dividends. Money started to move back into value, and the prior liquidity flywheel regime reversed with a vengeance. Suddenly value funds were posting great numbers and sidestepping major wipe-out tech sector losses, and so started getting accelerating inflows, and all that new money was plowed into low multiple stocks, and there were few sellers. Prices kept rallying and the cycle continued. This lead to a spectacularly good decade for value (and particularly EM value, where many markets went up >10x from extremely expressed and under-owned levels at the turn of the millennium, as money had previously rushed into the US tech/large-cap and the USD, and away from EM after the Asian Financial Crisis and a collapse in oil prices to US$15/bbl).

By circa 2010 (2012 for EM, 2014 for FM, and 2006-12 ish for DM value) we had come full circle. Value had outperformed over a full cycle with let risk, like it usually does (and without owning the 1990s cohort of secular winners, I might add), but multiple dispersion had now significantly decreased, to the point of being unusually low. I recall reading Grantham's quarterlies at the time, where he observed that multiple spreads were unusually low, and that in particular, that there was an unusually low multiple premium being awarded to super-high-quality in the US. 

What had actually happened over the prior decade was something of a bubble in value strategies (as bizarre as that might seem). On account of their strong performance over the past decade, too much money had flooded into traditional value strategies, which resulted in multiple differences being arbitraged down too much. It resulted in a number of extremely mediocre investors such as Whitney Tilson amassing large funds merely on account of being in the right place at the right time and getting lucky - buying low multiple stocks at the right time. They too fell into the trap of letting a liquidity fueled secular bull market in value cause them to mistake themselves for financial geniuses. As market conditions became more challenging for value post GFC, many have blown up, and in so doing, proved themselves to have merely been beneficiaries of favourable market conditions in the prior decade, rather than in possession of genuine investment skill (and also demonstrating that it can take up to 20 years to truly assess the capability of an investment manager - i.e. through a full secular market cycle). The value players with genuine skill are those that have invested through a full secular cycle and come out with good numbers, even if recent numbers have been poor (such as Platinum in Australia).

The pendulum has swung back the other direction since circa 2012 as a new wave of tech structural growth winners emerged from the wreckage of the bust and GFC. Valuations were more than sensible in 2012 - they were cheap. Google for instance traded at just 12x earnings ex cash in 2012. It was a bargain. Visa traded for just 15x. As Grantham accurately observed, high quality was very cheap. On account of their quality, growth, and low valuations, these stocks predictably did well - much better than value stocks which by then were expensive relative to high quality on a quality-adjusted basis. After the trauma of the GFC, and the low interest rates that emerged in its wake, investors also started to seek out safety and yield-substitutes, which triggered the beginning of a secular upswing in demand for high quality defensive equities, and particularly defensive growth. It made sense - for a while. 

However, at some point we veered down the same road we always do - valuations rose, driving strong returns. That has lead to performance chasing inflows, driving multiples up further, and so on and so forth. So here we are again, approaching the end (perhaps) of yet another swing of the secular pendulum, where virtually the entire world is long mega-cap tech and other software names, the USD, and defensive high quality (healthcare, staples etc) and defensive secular growth, with valuations near record highs; triple digit P/E ratios have lost their shock value (as have P/S ratios of >30x); all while other parts of the investment universe are generationally cheap with low single digit P/E ratios and 10% dividend yields - some of the greatest value dispersion we have ever seen in market history, period. 

In my view, value investing is not dead - it's just spent a decade going from fully priced to cheap while other market sectors have gone from fair-to-cheap to very expensive. The commonalities today with 1999 are very stark. Oil and energy is very depressed, just like it was in 1999 when The Economist published its famous 'drowning in oil' cover. Parts of EM are very depressed, including currencies, much as they were in 1999 after years of a strengthening USD and the fallout from the Asian Financial Crisis; value is discredited and value funds are shutting down, just like in 1999, and today, everyone owns the USD, tech, and US large cap quality stocks, just like in 1999. And in both cases, the boom has been aided and abetted by Fed monetary excesses. 

That said, we are arguably still are not yet at the level of valuation crazy we saw at the peaks of 2000 (although that's debatable), and we have far more monetary stimulus and lower rates now than we did back then (not debatable). As a result, it is always possible the bubble inflates to even more epic proportions for several more years, with stocks like Google and Facebook ending up at 60-80x earnings, although I wouldn't bet on it. On the other hand, the size of the aggregate market caps now in absolute terms are already much more extreme than in 2000. I'm not sure when a reversal will happen, and whether one is already underway, but in my view it is inevitable that the pendulum will swing back the other way at some point, because such forces are absolutely inherent to how financial markets work - as little appreciated as they often are. Always have been, and always will be in my view.