Scalper1 News
Summary “Smart Beta” and systematic investing strategies have become wildly popular in recent years. The trend has largely been driven by technological improvements and positive feedback loops. There are risks to systematic investing that must be acknowledged. Most importantly, systematic investors must acknowledge that stocks are not pieces of data, probabilities, or bets. They are legally, tangibly, and truly ownership interests in businesses. The Rise of Systematic Strategies According to Investopedia , “smart beta” was the most searched for financial term of 2015. Smart beta funds and ETFs are popping up all over the place. According to CNBC (emphasis mine) : As of June [2015], there were 444 strategic/smart beta ETFs in the market managing about $450 billion , according to Morningstar data. That’s up from 213 funds managing $132.5 billion in assets in 2009. They now account for 21 percent of all exchange-traded products and about 31 percent of all cash currently flowing into the industry . Anecdotally, fund companies like Gerstein Fisher (MUTF: GFMGX ) that have employed smart beta-like strategies for decades have suddenly seen a pouring in of assets. Before we go any further, what is smart beta? Investopedia defines it as follows: Smart beta defines a set of investment strategies that emphasize the use of alternative index construction rules to traditional market capitalization based indices. Smart beta emphasizes capturing investment factors or market inefficiencies in a rules-based and transparent way. The increased popularity of smart beta is linked to a desire for portfolio risk management and diversification along factor dimensions as well as seeking to enhance risk-adjusted returns above cap-weighted indices. It is a very general marketing term to describe (1) passive strategies (no active individual security selection) that (2) construct portfolios using weighting methods and metrics other than market capitalization weighting. Traditional indices are market weighted, and this has been observed to be detrimental to performance compared to equal weighting or fundamental-based weighting. By weighting, I mean the size of each position in the portfolio. An example of smart beta would be taking the 500 stocks in the S&P 500 (NYSEARCA: SPY ), but instead of assigning weights based on the market capitalizations (ex: Apple (NASDAQ: AAPL ) would be ~3% of the portfolio), you could weight the portfolio by LTM net income. For the purposes of this article, I’m more interested in smart beta for the general strategy and secular shift it represents – a shift toward systematic investment strategies that aren’t indexing, but aren’t individual security selection either. Outside of “smart beta” specifically, systematic strategies in general have become very popular. The success of Michael Covel’s Trend Following products and books, Tobias Carlisle’s The Acquirer’s Multiple product and books, Wesley Gray’s Alpha Architect , Joel Greenblatt’s Magic Formula and Gotham Funds (MUTF: GARIX ), etc. are evidence of this. What about the most popular investing blogs? Abnormal Returns , Pragmatic Capitalism , A Wealth of Common Sense . All these blogs have a systematic/passive bent. It seems to me that in the last year or two, systematic, rules-based strategies have become enormously popular. Maybe I didn’t have my eyes open before then, but now I can’t seem to avoid this stuff. Why? Technology. The rise is largely the result of technological improvements. Systematic strategies are fundamentally empirical. They require historical data and a backtest to answer the question “What’s worked in the past?” Technological improvements have made this possible. It’s now very easy to run a backtest on a Bloomberg terminal. More serious backtesters can use the extensive databases of Compustat and UChicago’s Center for Research in Security Prices (CRSP). And this feeds on itself, because people who do the research and backtests often publish their results, which are then used by other investors. So, more and more people have various answers to the question mentioned above and, naturally, more desire to do something with it. The other question: Is there a way we can run this strategy without human interference – fully automated? This is important because it’s difficult to manually follow a systematic strategy that involves purchasing hundreds of securities at potentially very short intervals, calculating weights, etc. It’s just not that feasible to do it manually. Technological improvements have made this feasible as well. I’m not so sure I understand the specifics of how this is done, but clearly, if hundreds of firms are doing it at much lower expense ratios than traditional actively-managed funds, there is automation involved. And this feeds on itself too. Once a fund/ETF has figured out how to do it, other investors can just buy into that ETF to participate in systematic investing. Personal Reflection This all is reflected in my recent articles and the evolution of my investment strategy. Being exposed to all of this has deeply influenced me. I also think, as I mentioned in a prior article, beginning to playing poker (a deeply probabilistic game) has had a significant impact. I’ve begun sourcing stocks using screens filtered by metrics that outperform, like EV/EBIT. I’ve begun taking small starter positions or bets, and looking at aggregate performance instead of performance by position. Put simply, I’ve begun to think of investments as bets and the future probabilistically. I’ve become empirical. Risks There is a lot of good in this transition, but I’m realizing now that it is dangerous if taken too far. Historical data is great, but there are risks to it. One is data mining, which I discussed in my article on stock screening. It’s worth googling “Butter in Bangladesh.” Then, there’s execution risk. What if your technology is flawed? What if there’s a power outage or you experience a data breach a la Target (NYSE: TGT )? What if you override the system at all? Joel Greenblatt points out that when he and his colleagues tried to source from Magic Formula without buying all the stocks on the list, the performance of the stocks they picked actually underperformed the market despite MF in aggregate outperforming, because they tended to avoid the biggest outperformers. They were the hairiest, and that’s why they performed so well. What if the markets change? The predictive power of metrics like P/B, which Fama and French articulated decades ago, has greatly diminished since. Past performance does not predict future performance. This is particularly important given the shift toward systematic strategies. The more popular these strategies get, the quicker the excess returns will be arbitraged away. Don’t assume you can stick to it either. It’s great looking at 50 years of data and seeing that over that period, the strategy has substantially outperformed the S&P, but that doesn’t mean there weren’t extended periods of substantial underperformance. In fact, most studies point out these spots of underperformance. One of my favorite quotes by Ben Carlson is this: The advice is to think and act for the long-term, which sounds great on paper, but the problem is that life isn’t lived in the long-term, it’s lived in the short-term… The problem is not the knowledge, it’s the behavior. Quitting smoking is not hard because people don’t know it’s bad for them, it’s hard because it’s habitual and it’s hard to change those bad habits. If you employ a systematic strategy, it’s because you think it will perform better than something else (most likely S&P 500) in terms of return, drawdown, etc. Naturally, you’ll be prone to comparing the performance of the strategy to that benchmark fairly frequently, and it will be difficult to see that it is performing worse over an extended period and still stick to it. To make this point more tangible, let’s use an example. You implement traditional Magic Formula (30 stocks, equal weight, annual rebalancing) with the expectation that your annual returns will substantially exceed the S&P 500. 4 years into implementation, you’ve underperformed in every single year (very possible) and cumulatively, the S&P 500 is up 15% annually and you are only up 8% annually. Unlike a fundamental research-driven active investor, you can’t explain this away with mistakes (“My current investment strategy works, I’ve just made mistakes and bad decisions along the way. My strategy is improved now and I’m more knowledgeable and experienced. I’ll do better going forward.”) The only thing you can do is question whether the selection criteria you are using still work. You only have four more years of data – data that disproves your initial hypothesis. That’s it. On top of that, clients and peers are badgering you about it. Surely, it’s difficult to stick to the strategy. Moreover, even if you want to stick to the strategy, there’s a good chance your clients don’t and they pull their money. At this point, you’ve stopped using the strategy at the worst time possible and managed to achieve underperformance with a strategy that has outperformed in the past and will likely outperform in the future. Stocks are Ownership Interests in Businesses I don’t mean to say that the empirical evidence is not compelling. It is. Some of these backtests encompass many decades and market cycles. Carlisle and Gray’s backtests in Quantitative Value are over 50 years. I also don’t mean to say that completely systematic strategies can’t work in practice. They can. The best example is probably Jim Simons’ Renaissance Technologies. The flagship Medallion fund did 72% annual returns before fees over a 20-year period from 1994 to 2014! What I am saying is that I don’t think a completely empirical approach to investing is sound, at least for me. There are too many things that can go wrong if we just leave it at this. Ultimately, stocks are ownership interests in businesses, not probabilities or bets. Maybe stocks can be thought of as probabilistic bets as a working assumption for a strategy, but that’s not what they actually are. A stock is legally, tangibly, and truly an ownership interest in a business. Ben Graham said this decades ago, and Buffett has singled it out as one of the 2-3 most important concepts to be learned from Graham. I think a much more sound approach to investing for empirically-driven, systematic investors is an upfront acknowledgement that goes something like this: Stocks are ownership interests in businesses. Stocks increase in price when the value of the underlying business increases or when there is a gap between the price of the stock and the value of the business and that gap closes. That is what is actually happening. As an investor, I have the opportunity to look at individual stocks and try to buy those whose prices do not fully reflect what the value of the underlying business is or will be. However, there is a wealth of data from historical markets that can be used to systematically identify these types of attractive situations. I feel, for various reasons, that these historical relationships are compelling and will continue to be. I also feel that I will be more successful as an investor using these systematic shortcuts than I would be if I tried to identify individual cases of undervaluation manually. The bottom line is that no matter who you are or how you invest, you need to acknowledge stocks for what they really are: ownership interests in businesses. Scalper1 News
Scalper1 News