Since the crash of 2008, many investment advisers and financial commentators pronounced that ‘buy and hold’ is dead. At the lowest point, the S&P500 fell as much as 50%, and even today most portfolios never recovered from the fall experienced by domestic and international markets, which both appeared to fall in sync. Worse yet, several high quality bond funds fell by a significant amount, prompting investors to abandon an asset generally regarded as being safe. These events led to the assertions that buy and hold is no longer a valid investment strategy. Active asset allocation was the new investment strategy, and being nimble as well as having the ability to predict which asset class will outperform in the future was the new skill to have. Is ‘buy and hold’ truly dead or are the practitioners to blame for their own failures?

Prior to the crash of 2000, it was generally believed that the best way to get market-beating returns was to hold a handful of well-known dividend paying stocks with good balance sheets and a high growth potential. To find the right stocks, one had to study the balance sheets and financial information as well as to evaluate numerous measures based on company financials. Once the work was done, buying and holding was the key to success. If the stocks fell, the advice was to buy the dips. Unfortunately, this strategy fell apart just as the market started crashing in early 2000. Most of these stocks fell 50% or more, and some fell even lower. Many companies failed or got bought out. In hindsight, it was suggested that the price to earnings ratios were too high, and that the market was grossly overvalued, but at the highest point almost every single company’s multiples were higher than historical average, though very few experts suggested that this meant to stay out of the market altogether. Nothing prepared advisers for such a calamity, and in a desperate effort to get out from under the falling markets, many cashed out their portfolio, just like countless other investors. Many of the stocks never recovered in the following decade, and some went even lower. So, ‘buy and hold’ seemed to have failed.

Or did it? Several investment managers were able to make it through the storm while holding a globally diversified portfolios of mutual funds. Their claim was that all one needed to do was to design a portfolio using Modern Portfolio Theory (or MPT), and then all would be well. Was MPT the magic recipe for portfolio management? I set out to learn the secrets of MPT by studying the course material of several openly available MIT graduate level courses on investment management. It was perfectly logical that MPT proponents would use ‘buy and hold’ as their investment philosophy. The MPT states that diversification can be attained by holding a basket of uncorrelated asset classes. According to MPT, diversification provided risk management which would in theory be able to soften the crashes like the one that happened in 2000. Not only does MPT provide risk management, but it also provides for excess returns (or returns higher than those of a risk-free asset such as treasuries) based on historical returns of each of the underlying asset class. This sounded exactly what I was looking for. No more worrying about how to select individual stocks, and a method to manage portfolio risk was a definite bonus. This meant getting into hundreds of stocks, but it could easily be accomplished with mutual funds. Several Nobel Prizes were awarded for the MPT, and it was well developed mathematically, with many theorems and proofs that were relatively easy to follow. Everything was consistent, and seemingly bulletproof. Backtesting showed that diversification worked very well during the crash of 2000, and a properly diversified portfolio held steady, barely losing anything during the time when most people lost their shirts. This approach called for a fairly static allocation that changed with age, and the allocation to stocks was based on risk tolerance of the individual investor. For most advisers, MPT was like gospel, and hardly anyone questioned its validity prior to the crash of 2008.

Not everybody in the investment industry was convinced that MPT worked. The world of investments changes rapidly, yet it appeared that MPT was never seriously challenged since it came about in 1950s. For a theory that appeared to have such staunch backers, even mild criticism was something worth investigating, so I began reading the literature critical of MPT to see if there was anything behind it. The year was 2006, the stock market was going up, and no worries appeared on the horizon. The logic of MPT practitioners went like this. Do you have a high risk tolerance (and many people do when the stocks are going up)? No problem! Put all of your money into a diversified portfolio of stocks, and forget the bonds. Who needs bonds anyway? Bonds don’t grow much, yield little, and are nothing more than ballast. Besides, the industry pays a lot more to sell the stock funds than the bond funds. Just double up on stocks – diversification will do its job, just like it always did. Everybody was using MPT, and even institutions used it to gauge risks of their own portfolios. But something didn’t add up. MPT is based entirely on Normal (or Gaussian) distribution known as the Bell Curve, which implies that large market moves are very rare, and extremely large market moves can not ever happen. The only problem with this was that these moves not only happened, but they happened all the time and their effect was exactly the force that was crushing portfolios left and right. But in 2000, many diversified portfolios survived, so nothing could have prepared the investors for the crash of 2008, when every portfolio, regardless of how diversified, plunged with the markets. After years of such market turbulence, which became progressively worse after 2000, even the most seasoned investment advisers all but gave up on ‘buy and hold’ and MPT. But was MPT to blame? Upon a closer look, it appears that the entire theory of MPT is based on several assumptions which are not always true. So what, one could ask? The answer to that question lies in an obscure paper of Benoit Mandelbrot.

I’ve always known Benoit Mandelbrot as the mathematician who popularized fractals. However, in 1962 Mandelbrot published a seminal paper that had long remained forgotten primarily because it was at odds with the views of investment establishment. Only relatively recently the points Mandelbrot raised as far back as 1962 started to be taken seriously. In “Variation of Certain Speculative Prices” Mandelbrot has shown that cotton prices behave nothing like the Normal distribution would predict, and instead they exhibit highly chaotic and much more extreme behavior consistent with another distribution – what is called Stable Paretian Law, which comes from a family of scalable power laws. Over the next several decades, Mandelbrot expanded this work into a theory called Multifractal Model of Asset Returns. Without getting into the details, the implication of this discovery would deliver the first crack in the foundation of the MPT. If the world was less Normal and more Stable Paretian, MPT did not hold. A statistician from MIT’s Sloan School Paul Cootner wrote in 1964:

“Mandelbrot, like Prime Minister Churchill before him, promises us not utopia but blood, sweat, toil and tears. If he is right, almost all of our statistical tools are obsolete . . . Surely, before consigning centuries of work to the ash pile, we should like to have some assurance that all our work is truly useless.”

Having discovered Mandelbrot’s contribution to finance, it wasn’t long before I found Nassim Taleb, the famous commentator and the author of “The Black Swan” who considers himself to be Mandelbrot’s disciple. Eventually it became clear that the missing link in the ‘buy and hold’ concept was proper risk management – not the traditional understanding of risk as dictated by the Modern Portfolio Theory, but the risk management paradigm that comes from understanding the nature of Taleb’s Black Swans and Mandelbrot’s Mild and Wild randomness, with the two extremes described by the Normal distribution (Mild) and scalable power laws (Wild). Thus, MPT may be ‘correct’ most of the time, but when it fails, the failure is so spectacular that it is justified to consider MPT a total failure. As Nassim Taleb put it, “It does not matter how frequently something succeeds if failure is too costly to bear.”

There is one big issue with Mandelbrot’s mathematics as far as practical application is concerned. There are neither tools nor easy to implement formulas which can make investors lots of money, at least not yet. The math simply tells us that we underestimate by orders of magnitude the real risk of investing in the markets. Thus we are left with more questions than answers. To solve this dilemma, Taleb recommends using the ‘negative knowledge’ approach instead of using a prescriptive approach that is generally practiced. Instead of claiming knowledge without having any idea as to the confidence of that knowledge, Taleb recommends simply avoiding the areas where our knowledge is lacking substantially, especially when dealing with the products and markets where the risks are routinely underestimated and where failure can be costly. After we admit that we can not estimate risks with any degree of confidence, Taleb’s recommended approach is to be as conservative as possible with most of your portfolio, while taking wild risks with a small portion of it. Taleb recommends having as much as 90% of your money in CDs or guaranteed securities (such as Treasuries) while keeping 10% in high risk out of the money options, which are very often mispriced because of general misunderstanding of risks by the financial establishment. While this may work for a select group of high net worth individuals, somebody whose portfolio is mostly concentrated in a 401k will probably not be able to implement this approach. A similar approach can be implemented in a 401k without taking nearly as much risk in the ‘risky’ portion of the portfolio. The key is to disregard the traditional advice about allocation to stocks vs. bonds, and to allocate much less to stocks and much more bonds, as well as to select your investments carefully with an eye towards hidden risks and transparency.

Now that we know that markets are turbulent and that risk is severely underestimated, what are the implications for the traditional ‘buy and hold’ approach, and what was wrong with the ‘naive’ approach to ‘buy and hold’ practiced by most MPT believers? The lesson is that holding a small number of stocks does not lead to diversification. A portfolio of 1000 stocks is also not diversified when invested in a single sector or country. Even holding every stock in the world will not protect your portfolio from large and sudden crashes. Sometimes diversification works, and sometimes it does not. But when it fails, the failure can be spectacular. A conservative balanced model portfolio (with 50% invested in fixed income) that had a small gain from 2000 to 2002 when most investors were losing their shirts had fallen as much as 25% when the S&P500 fell 50% in 2009. While this may be a relatively good result (the return for the year was a modest -8%), if you had millions invested in this portfolio would not have been happy right after the plunge. The markets later recovered, but the lesson from this is that any investment in stocks is riskier than most investors realize. Because the losses are sudden and large, it only takes a single big loss to derail the plans of many investors. It may never be possible to bring that money back, since it could take more time than investors have available, and there is no guarantee that future returns will make them whole again. The conclusion from this is to have no more than a fraction of your money invested in stocks at any one time. This may sound as overly conservative, but it only takes one crash to lose a big chunk of your portfolio. If this happens when you have already retired and are not contributing any more to your 401(k), you may be in trouble.

The lesson for the ‘buy and hold’ investors are clear, thanks to Mandelbrot and Taleb. Diversify as widely as possible and always be more conservative than the experts recommend simply because the markets are much more turbulent than is traditionally believed. ‘Buy and hold’ does not guarantee a good result by itself – the recipe is to control portfolio volatility by limiting exposure to asset classes whose behavior can at times be chaotic. Though it is counter-intuitive, minimizing portfolio volatility will lead to more consistent returns in the long run given the realities of the markets, and while we still don’t know much about how the markets work, sticking to what we do know and limiting our exposure to the effects of what we don’t know is a worthwhile risk management strategy.