This fourth piece focuses on a different way to view the goals of an investments firm. It is my theory. I do not think I have provided enough evidence to support some of my claims, there are some parts that are still bugging me. I’m trying to find a more useful way to theoretically evaluate active investment firms, instead of resorting to alpha. In my experience, particularly after learning the way alpha is calculated, it is often a weak measure of performance. Using alpha assumes that the firm isn’t doing useful research in various beta factors, which could arguably be more useful for their clients. Simply beating peers in an incredibly uneven market does not mean much. Instead I argue that infrastructure, talent, and research demands positive economic rents. And this is more fruitful of an approach for self-analysis as well as evaluation of a fund. In the end the alpha of an entire market is zero, but the value created by intelligent investors evaluating equity, debt, and other products is decidedly positive. This is my attempt to solve the problem of “how does active investing make sense if alpha sums to zero and markets are efficient?”
Ricardo’s theory of Rents:
I believe Ricardo’s theory of rents is applicable, yet overlooked, when considering the investments industry. It is a fundamental piece of economic theory that was created by one of the founders of microeconomics. The theory of rents helps destroy the dichotomy between alpha and beta. While alpha and beta are useful for constructing a model of reality, it is an attempt to simplify the world to an equation. There are some serious flaws with the alpha and beta model. The first is that all alpha must equal zero since every investor that beats the market takes from someone who lost.
Often academic research suggests that active investing is a pointless game due to the fact that all alpha across the board must equal zero by definition. As I have mentioned earlier, this is due to alpha being a measurement of return without risk or beating the market. And for every winner there must be a loser, resulting in alpha amounting to zero over the whole market. This is the definition and theory behind alpha measurements, however the reality of measuring an entire market to zero-out the alpha is impossible.
Viewing the primary goal in a market as only alpha returns is also problematic, because this suggests the market is a zero-sum game. However, if all active investments switched to passive investing (not taking active bets) then the market would not change to reflect new information. This would cause the markets to become inefficient, which suggests active investing is not a zero-sum game.
Moreover, the widely accepted view of economics is that markets tend to be efficient under a capitalistic system. While academia does often identify areas of inefficiency in various fields, it is suspect to claim that most active investments firms are actually not contributing to social welfare and utility. As a result it is important to search for a methodology that allows for the existence of active investments to make sense, while not requiring that the goal of active investment firms be to generate risk adjusted returns (alpha).
Imagine that we reject the assumption that the goal of active investing should be to return alpha. If this were the case we would ignore residual values in regression returns (which is how we identify alpha) and only concern ourselves with risk factors (beta). I posit that active trading would still be beneficial. Some risk factors are not open to easy investments. For example, maintaining exposure to inflation risk on an emerging market junk bond portfolio. This would require a hedge fund with talent, infrastructure, and a great team. This firm would have to charge money to support its firm, and it might be one of only a few hedge funds to offer pure exposure to this risk factor (as opposed to holding small amounts of it inadvertently when buying a market index). This hedge fund would be accessible only to high net worth individuals, would charge expensive fees, and would provide potentially high returns with low correlation to traditional investments. This is a useful service that investors wish to buy, and it does not require this fund to beat other investors.
Contrary to previous examples, I have described this using only risk factors. By using Ricardo’s theory of economic rents, we can identify active investments firms that provide a useful service without obsessing purely over the residual value (alpha) of regressions on their investments and/or whether or not they beat their peers. Hopefully through explaining this theory and then revisiting an example I can convince the reader that obsession with alpha is overrated. If you are not familiar with Ricardo’s theory of rents I have included a detailed description at the end of this piece.
A primer on economic rents can be seen by viewing a recent high profile case in contemporary finance: Groupon (GRPN). This firm added a new dimension to deals and coupons. They turned the coupon industry into a fun way to try new things targeted at young social media users. And they flourished at first. However, their stock has dropped from $25 to $4 in six months. One reason this happened was due to a decrease in Groupon’s economic rents. They had a first mover advantage, since they essentially created a new product. In this short-run Groupon made immense profits. These profits were far and above what a firm ought to be compensated for the act of organizing and emailing coupons. Since there were no competitors at the time they were able to capture massive rents. Eventually the field was swamped by imitators, which drove down Groupon’s economic rents to zero. In fact, so many investors thought this would be happen, Groupon was briefly the most shorted stock in the US market. This means owners of Groupon shares could charge a very high rate to speculative investors who wanted to borrow the stock from shareholders so that the speculative investors could sell it short.
Understanding the idea of economic rents is fundamental to economics and financial markets. Economic rents are zero in all perfectly competitive industries. Even though these industries make money and profits, their economic rents are zero. As I explained earlier this seems counter-intuitive at first, yet makes perfect sense. It takes constant innovation above and beyond the market norm (or a monopoly) to keep rents positive. A simple innovative shock will temporarily bump up economic rents but then decrease to zero once again as other firms copy. For example, Apple was able to increase its economic rents recently by winning the lawsuit claiming that Samsung infringed upon their innovation. The courts decided Apple deserved economic rents due to innovation that Samsung allegedly stole.
Determining economic rents is difficult and often results in debate, as seen in the Apple and Samsung lawsuit. It is even more difficult in the investments industry. It is a more theoretically complex view than simply measuring the alpha residual on risk adjusted returns. Instead we view these firms to determine how they are providing services based on their inputs. For example, the currency hedge fund I hypothesized at the beginning might have zero or an immeasurable ‘traditional’ alpha. Despite this, the firm would be operating with positive economic rents due to their ability to bring exposure to a complicated risk factor to their clients, a service that is not easily replicable by other firms.
Another example might be a fund that holds the equity of many large Chinese firms in their portfolio. In contrast to other emerging market investments firms, this fund has many teams of investments professionals that worked at all the major Chinese firms before they joined the fund. As a result this fund invests in select groups of Chinese firms with a decidedly above average ability at picking stocks. A traditional risk-adjusted return model, such as the Fama and French three factor analysis, would measure their investments against a few well documented beta values and attribute the excess residual to alpha. Yet this analysis falls short. Consider instead that the team of investors sat together and thought of these different firms, analyzed their balance sheets, and viewed massive amounts of data. After this analysis they determined these Chinese firms would provide strong returns. Instead of measuring this basket of stocks against traditional market risk factors, we could argue that this team created a set of risk factors for each stock using macroeconomic data and the population as factors. For example, an agricultural firm might have its pork returns tied to the growth of GDP per capita in developing cities. In this example the team researched and chose a portfolio of stocks. It is possible the firm earned no traditional alpha. Instead, they identified factors that they believe will be compensated strongly based on their prediction of the future. And as mentioned earlier, predictability does not imply market inefficiency.
I argue that viewing this past example in terms of positive economic rents has more use than attempting to search for alpha. First let’s consider how using a traditional alpha calculation, similar to examples I offered in the past post, would work. Let’s suppose that this investments firm uses a value oriented approach, and searched for ten Chinese stocks that they believe traded at a substantial discount. A rudimentary (yet typical) risk-adjustment would look for what amount of the returns can be explained by traditional factors. This would begin by identifying a market index so that the market beta value can be ‘taken out.’ Additional factors such as Fama and French might be adjusted for as well. Every aspect of these ten stocks that is not explained by these factors will be lumped together in the alpha category. This alpha category while containing some information, is far from a good indicator as to the success or failure of the firm.
Imagine that the bulk of the research from this investments firm was to find firms that created goods that have dramatically increased in inelasticity over the past decade. Perhaps they were searching for goods that were once considered luxury before the rapid economic growth. As a result this investments firm buys these stocks in the hope that they will continue to deliver consistent returns even in the face of a recession. Their conclusions was that the market had only partially priced in the defensive nature of these firms. As a result they believe that over the next half decade or in the next recession, other investors will realize this defensive quality is worth paying extra. Simply put, this firm wanted to buy a basket of defensive Chinese stocks, based on their own rigorous research. Based on this hypothesis, if the following year still has strong growth in China, the investment firm will have a negative alpha value. In theory, the firm might argue that if this ‘defensive beta’ factor they believe existed were included, their lower returns would be justified. However, they could only argue this theoretically. And this is why traditional alpha measurements can be flawed and misleading so often. The theory of the law of one price, risk and reward, factors, and alpha is essential to understand. The proofs that create this framework teach us how markets operate beyond what we are able to truly and scientifically measure.
It is for this reason many academics argue that stock picking is impossible. What they really ought to say is that we have yet to measure stock pickers generating true alpha values in a manner consistent with the scientific method. And many investors, while unable to prove what they have done is true skill with measurements that qualify at the peer reviewed journal level, are still able to make money off of their investments. Yet positive alpha values are not necessary to justify the existence of stock pickers. After all, alpha is zero in the market as a whole. Consider the previous example. The active investments firm that was stock picking in China was not searching for positive risk-adjusted returns. Instead they were searching for a factor that they believed would perform well over the following decade. Similarly, stock picking can exist even without markets being inefficient and providing alpha values. Uncovering the right beta values in the right combination is difficult, requires research, and can be rewarded. My example of ‘defensive Chinese’ stocks was still overly simple, as it only claimed there was one risk factor common in all ten stocks. The reality of stock picking is a firm might have a portfolio of thirty stocks. And each stock might have ten different factors that the researchers argue will allow it to perform strongly. The term factor might even lose its meaning, as some of the researchers make their investment based on a firm handshake with the CEO.
While the law of one price claims similar risk must contain similar reward, these risk factors might have low correlation with traditional market returns. If this is the case, it is possible the fund managers found a way to gain their clients exposure to a subsection of the market that will either improve their returns and keep their volatility constant or keep their returns constant and lower their volatility.
Consider the following example: Ten hedge funds all are predicting the outcome of a single firm over the next decade. Each one believes this firm is exposed to a few rare factors that suggest it will provide consistent returns. Each fund engaged buys the firm, and receives the returns. Depending on when each firm bought or sold this stock, there will be a group of winners and losers relative to one another if we only rate them based on their alpha value. However, each fund might have bought the firm for a different reason, to satisfy a different portfolio, and off of their own independent research. Perhaps the fud that received the lowest return per risk ratio bought this firm based on research that suggested to them the times when a firm is at extremely low risk for a defensive portfolio. While it may appear that they came last in the group of ten, they might have actually identified a period of calm. And while they were rewarded less for their risk, as we would expect in an efficient market, the firm still satisfied the goals of their portfolio.
My conclusions is that using Ricardo’s theory of economic rents might better explain the advent of active investing and stock picking, as opposed to searching purely for alpha. And as always, in markets with a large variance, some investors will by nature do better than others for reasons that we tend to call luck. This viewpoint also allows for active investing to make economic sense. Large amounts of academic research, after finding that alpha does not beat infrastructure costs, would appear to suggest that many active investment firms are not meeting their goals. However, this is problematic for the same reason I argued in the prior example with the Chinese firm that was searching for a new and ‘rare’ factor. I believe instead we may view active investments firms as firms that are rewarded for their economic discovery of various market, economy, and individual stock, factors. While the true alpha of an investments firm might be zero, it is possible that the firm has positive economic rents. They would be awarded these positive economic rents for their research and innovation in identifying, studying, and investing in various different factors that might affect the price of stocks. Such as the affect weather will have on oil shipments or a rising demographic of tweenagers that are more likely to buy Apple devices. It is possible that these firms make money, generate positive economic rents, and do all of this by identifying the right mix of the infinite level of factors and even discovering new factors. Similar to my original currency example, it is possible that a risk factor exists, and is compensated, yet does not exist as a pure investments vehicle. If a firm is able to research this factor, and isolate it, investors will pay them to gain and maintain exposure to this factor using their infrastructure capital investment and human capital and human labor. In doing this the firm will increase societal wealth by allowing a more efficient transfer of risk and create positive economic rents. All of this can be done successfully without focusing only on beating their alleged competition, so as to generate alpha returns. Instead investors can be viewed similar to other professionals. The work of an intelligent investor in identifying risks in the market is compensated. The industry practice of tethering this compensation to a figure that is measured ‘easily,’ such as alpha makes sense, as it gives an investor the incentive to work hard. However, this does not suggest that alpha is the true end all goal for a portfolio manager.
Ricardo Rent Explanation:
I will begin with an example, which involves two nearly identical hot-dog stands. Hot dog stand A is in New York City and hot dog stand B is in a small town called Springfield. Each owner needs to buy a business license for a mobile cart from the city to operate, and each city charges a price set by the market equilibrium of supply and demand. In this scenario let us assume that the functions that restrict supply and create demand are only based on the population of each city. Now, let’s guess how much each hot-dog stand owner is paid. Each stand owner is doing the same job at its very core: Managing a stand, cooking, and selling hot dogs. However, the New York stand should sell more hot dogs simply because it is in a higher population city (this is a key observation). Now let us consider what the cost for a food cart license is in each city: We know each city issues license proportional to the population. It is likely that the gut instinct of the reader is that the New York license ought to cost more, this does make sense based on experience. But why should it cost more? Well, each city that issues a license wants a portion of the rents that they deserve for having the ability to grant the right to sell to consumers. The right to sell to all of New York is greater than the right to sell of all of Springfield. As a result the New York license will be cost more than the Springfield license. The cost should be enough that the equilibrium wage of each hot dog stand owner is identical (reality will of course deviate, but the general theory is reasonable, fast food workers tend not to have radically different qualities of life).
The beauty of this theory is that despite other factors, the employee of a hot dog stand only receives money for his or her marginal contribution to serving hot dogs. In this example one hot dog stand owner was able to sell far more hot dogs at a higher price due to a large and jam-packed city. But all that extra money, or rents, was demanded by the city when issuing the license. The city implicitly states that all rents captured due to the large population of the city do not belong to you, and the city demands them back when issuing the license. This will all happen in a competitive market in equilibrium. Ultimately consider that the two hot dog stands are essentially identical, each man pulls the same cart, cleans it, cooks hot-dogs, and exchanges them for money. Both of their economic rents are equal to zero, so ultimately they are only rewarded for their raw inputs and actions. This is economically intuitive, as we do not expect a fast-food worker in a large city to make far more than one in a small city. By recognizing that their economic rents are zero, we are able to acknowledge that while their revenue and sales numbers will differ drastically, they still ought to take home the same amount of money, as suppliers (in this case we are only considering a city licensing bureau) are able to demand higher shares of their revenue and sales.