LONDON (Reuters Breakingviews) - The U.S. stock market has ventured into bubble territory. Bond yields across the developed world are priced to deliver negative real returns. To make matters worse, the 40-year bull market in fixed income securities appears to be coming to an end. Rising long-term interest rates pose a serious threat to investors’ wealth. Their best protection may come, not from the likes of Alphabet and Amazon.com, which performed strongly last year, but from the antithesis to these high-fliers. Value investing is poised for a comeback. The investment approach has a distinguished pedigree that can be traced back to Warren Buffett’s mentor Benjamin Graham in the 1930s. There was a time when stocks that were cheap relative to earnings, book value or sales were more or less guaranteed to deliver superior returns. Between 1926 and 2009, American value stocks outperformed more expensive “growth” stocks by over 4% a year. Compounded over more than three-quarters of a century, that makes for a lot of outperformance. But the subprime mortgage crisis was particularly unkind to these stocks, which at the time were concentrated in vulnerable sectors, such as banks and homebuilders. In recent years, value shares have been crushed again. In a new paper for the Financial Analysts Journal, Robert Arnott and colleagues point out that the sector underperformed growth by 55% between 2007 and mid-2020. This was the greatest and longest drawdown suffered by the investment strategy since records began. Different reasons, some better than others, have been advanced to explain why value has done so badly. It is widely believed that stocks which trade on lower earnings multiples do less well when interest rates fall. In technical jargon, they are said to have lower “duration” than growth stocks. But Arnott finds that value’s recent underperformance is not correlated with the decline in interest rates. Furthermore, Ben Inker, the head of Asset Allocation at the Boston money manager GMO (and my former boss), argues that, adjusted for portfolio turnover, value and growth strategies have similar durations. In the past, the outperformance of value shares came in large measure from their profitability improving over time. It has been posited that the advance of the internet and the recent plague of corporate “zombies” have interrupted the process of mean reversion that historically benefited them. But there’s little evidence for these claims. In fact, value continued to do well in Japan throughout the 1990s at a time when zombies proliferated and interest rates fell to zero. The simple truth is that value stocks have performed poorly because investors chose to pay a heftier premium for growth stocks, and, in particular, for a handful of large technology stocks. The strategy has also been hurt by the trend towards passive investing. Vast inflows into capitalisation-weighted index funds benefited the so-called FAANG stocks (Facebook, Amazon, Apple, Netflix and Google parent Alphabet) while sucking investment capital from value strategies. Retail investors trading on Robinhood and other no-fee trading platforms have added to the noise, driving up the share prices of Tesla, Palantir Technologies and sundry special purpose acquisition companies. Value investing never does well during bubbles. To cap it all, downtrodden stocks, which are often found in more cyclical sectors, suffered badly during the pandemic. Big Tech, on the other hand, thrived from the enforced screen-time of lockdown. By last August, growth stocks were trading at an unprecedented 13 times the book multiple of their value brethren. Some of this rerating was deserved. Growth companies have recently shown an ability to maintain their pace of profit growth for longer than in the past. But financial markets tend to take things too far. Arnott’s colleagues at Research Affiliates, a Newport Beach, California-based investment adviser, find that value reliably outperforms when stock market bubbles burst. The last time U.S. stocks were this expensive was at the peak of the dot-com bubble at the turn of the century. After that bubble deflated, value blazed a trail of outperformance that lasted for several years. Value also does well in the early stages of an economic recovery. Market sentiment is turning back towards this out-of-favour investment style. The underperformance of value shares ended towards the end of last year as investors began to contemplate an end to the pandemic. In November, the Morningstar index of U.S. small-cap value stocks climbed by over 20% and added another 11% in February. International value has also rebounded. Even after these gains, Arnott argues that the sector remains relatively cheap, and has the potential for further outperformance as the global economy recovers. There’s a catch. Value stocks in the United States may be less overpriced than Tesla and Netflix, but they aren’t particularly cheap relative to their own history. What this means is that when the S&P 500 Index sells off, these shares on average are likely to lose money, even if they do better than the overall market. The good news is that there are better opportunities elsewhere. In the United Kingdom, value stocks were first crushed by Brexit fears and later hammered by a string of lockdowns. The outlook is brightening, however. Britain has already vaccinated a large share of the population and its departure from the European Union, although bumpy, has not been as disastrous as doomsayers predicted. Research Affiliates believes that UK value offers the “deal of the decade”. Value stocks in emerging markets have roughly the same expected returns as their UK counterparts but come with more volatility. Both Arnott and Inker agree that Japanese small-cap value is particularly attractive. A diversified basket of international value stocks can be expected to deliver annual real returns of around 5%-6% over the coming years. That’s a fair, if not great, bargain in absolute terms and perhaps the best chance investors have of preserving capital when the next market downturn arrives, as it inevitably will.