Is risk rational?
Misunderstanding of risk was a major factor in the subprime crisis and ensuing recession. Andrew Lo argues that one has to look at both logical and emotional parts of the brain to grasp how people respond to financial risk.
Q: What is risk? It seems like a multifaceted term, ranging from how I decide to drive down the highway to how hedge funds are related to banks.
In order to fully understand what's been going on in the financial crisis, we need to go back to basics and define what we mean by risk. For individuals, risk is anything that they perceive as threatening their well-being. Threats come in many different forms, so that risk is, in fact, a multidimensional attribute. We have to remember that fact as we start considering economic contexts, because individuals very often will confuse risks to their financial wealth with risks to their physical health. That's perfectly natural, because the decision-making mechanisms that we have developed over thousands of years of evolution don't always distinguish between those two. If we begin with that definition, we can start to see how it is that behavior ends up going off the rails in so many different contexts.
Q: What kinds of emotional responses do people have to risk, and when do they become inappropriate in a financial context?
The first point to make is that humans have very elaborate neural systems to address threats. We didn't get this far by being passive with respect to our environment. But there is a very important distinction between threats in the modern world versus threats on the plains of the African savanna at the time we branched off from chimpanzees and other great apes. In particular, the emotional circuitry that all of us have as part of our so-called mammalian brains kicks in to protect us from perceived threats. We've all heard of the "fight-or-flight" response. That's probably the most significant kind of emotional circuitry to be activated when we are threatened. Unfortunately, when we are faced with threats from the modern world, such as financial threats, that same circuitry kicks in — increased heart rate, blood pressure, and so forth. While that response may be very helpful in protecting us from physical threats, it's actually counterproductive in protecting us from financial threats, which are much more subtle and require the use of different parts of the brain, parts that shut down in the face of extreme emotional stimulus.
Q: Is emotion a clouding factor when you're thinking about investing?
Yes, extreme emotion is. Recent research in the neurosciences has greatly enhanced our understanding of the role of emotion. For example, Antonio Damasio at USC has reached the rather startling conclusion that emotion is absolutely essential for what we consider to be rational behavior. In his wonderful book Descartes' Error, he describes patients who have undergone surgeries to remove tumors in regions of the brain responsible for emotion, and afterwards these individuals act very irrationally, focusing on minute and irrelevant tasks for hours on end without any concern for missed deadlines, waiting clients, and other responsibilities.
You need a certain degree of emotion to just get up in the morning and get things done. You're going to write this article for the magazine because there's a deadline and because you want to achieve something, and if you don't, you will be suitably embarrassed, frustrated, and stressed out. All of those reactions are emotional responses that help you to get things accomplished in your life. Imagine if all of those emotions suddenly disappeared.
But with too much emotion, we very quickly become overwhelmed and the ancient circuitry that protects our physical existence kicks in. The fight-or-flight response is probably not an appropriate reaction to your deadline; you don't want to run away from that responsibility, nor do you want to kill your editor for imposing that deadline. And yet when we are overwhelmed by extreme emotional stimulus — for example, losing half of your 401(k) plan over the course of three months last year — that can be extraordinarily traumatizing. The fight-or-flight response may very well have kicked in for a large number of people around the world. And that will lead to potentially unproductive responses.
Q: It seems like self-interest, which is the basis of the rational model, is also at the heart of your emotional response, right?
Exactly. But when emotion gets out of balance with the logical faculties of our brains, that's when behavior becomes irrational.
Q: And the rational model of economics doesn't necessarily take that into account.
The rational framework of neoclassical economics and the efficient markets hypothesis are not completely wrong; they're simply incomplete. They focus on the behavior of markets and economic agents during "normal" times, during times when there is that proper balance between emotion and logical deliberation. What the behavioral economists have focused on is just the opposite — circumstances where that balance is out of whack and people act quite irrationally.
In fact, neither of these schools of thought is correct; they are both different halves of the complete picture, which I call the "Adaptive Markets Hypothesis." In that framework, individuals are simply biological agents that end up reacting, adapting, mutating, competing, and evolving. Instead of focusing on either of the two extremes of rationality and irrationality, we should acknowledge both aspects of our behavior and try to understand how and why we switch from one mode to the other.
Q: In one article, you compare a disaster that befell a mountain climbing expedition on Mt. Hood to what has happened in the financial system. Could you explain that?
That was a situation where some very experienced climbers made what many considered a rookie mistake. And the reason was because their perception of risk was actually quite a bit lower than the reality. We all have mental models of the world — all of us, except perhaps a few fortunate Zen masters. In many cases, those mental models are misleading. Some of us think of ourselves as better athletes than we are, or as more attractive than we may be. It's important to bring these models into agreement with reality when discrepancies can actually hurt us. Ultimately we do that through the forces of natural selection. We learn by trial and error. In the case of the mountain climbers, unfortunately, the error was extraordinarily costly in terms of taking their lives.
Q: And you see a similar dynamic in the build-up to the financial crisis.
Absolutely. The example of the mountain climbers is so instructive. Many people are now asking, "How could it be that some of the most sophisticated financial institutions in the world ended up making what now, in retrospect, seem like rookie mistakes, such as taking on too much leverage?" But the fact is, when you're building a business and the perceived risks have declined because of years of success and prosperity, well, you become complacent about some of those risks in the same way that these very experienced mountain climbers figured this was not a very challenging mountain. It was that false sense of security and perceived lack of risk that led to their demise.
This is not just a metaphor. This is exactly what happened with those financial market participants who took on too much risk, who over-leveraged, who became too aggressive about building their businesses. They didn't think it was all that risky. They actually thought that they had been able to deal with those risks reasonably well. Unfortunately they were wrong.
Q: Can I push the comparison one step further? The climbers were tied to each other, and then they were right above another group, right? All the ties between people and institutions seem important.
Yes, we are all "tied" to each other. In fact, there is an interesting although somewhat morbid phrase that the mountain climbers use: When you tie yourself to another climber and that other climber doesn't engage in "belaying," that is, securing himself to the side of the mountain using pitons, screws, and other devices, then you've entered into a "suicide pact" with the other climber — your fate is literally tied to the other climber's fate. And, in effect, all of us were tied to the U.S. residen-tial housing market. We just didn't know it.
Q: Is it much easier to overlook the risk in that sort of tie than one that is a threat to our physical well-being?
The financial world that we live in is so much more complex than even 10 or 15 years ago. Fifteen years ago, the word "subprime" meant sub-standard beef. The modern world is fraught with dangers that our biology is not adapted to address naturally. So we have to work at it. We have to engage in research and development and create technologies to protect ourselves from these kinds of dangers — in the same way that mountain climbers use different technologies like pitons, crampons, and synthetic ropes to protect themselves.
In some of the research I've done over these last few years on hedge fund risks, I've documented a number of new sources of risk that didn't exist 15 or 20 years ago, largely be-cause of the tremendous influx of assets into various exotic trading strategies. One way to deal with these complexities is to design more sophisticated tools, in the same way that we developed the wheel, the lever, and all of the tools that are responsible for the wonderful prosperity that we enjoy in modern society.
Q: Do you think you're close to having improved tools?
Absolutely. In fact, we've had them over the last five years. In papers that I wrote in 2004 and 2005, my co-authors and I warned about the impending crisis, because we actually saw it in the data using some of these new tools. The New York Times wrote an article in September of 2005 where they pointed out that the hedge fund industry was heading for a major dislocation, according to our research. And it was ignored, because, as the former CEO of Citigroup, Chuck Prince, said, if the music is playing, you've got to dance. So people continued dancing. Until the music stopped.
Q: You run a hedge fund. How do you handle risk from the practical end, as opposed to the theoretical end?
I think that the tools that we developed are, in fact, quite practical. So we practice what we preach. And so far it has held up well, precisely because we manage those risks aggressively and dynamically. That's one of the most exciting aspects of academic finance — some of the most esoteric analytics have immediate practical applications.
Q: What are some of your ideas for improving how the whole industry responds to risk?
The first point that I've made in some of my writings is that the financial crisis is really not about losing money. People have been losing money ever since there has been money. The real problem is when the wrong parties lose money, and by that I mean parties that are not properly prepared for those losses. For example, when a hedge fund investor loses 50% of her investments, that's unfortunate, but it's hardly a national tragedy. On the other hand, when a money market fund, a bank, or an insurance company loses 50% of its assets, that's a major problem, because those institutions (and their clients) are not prepared for those losses. And it must be the case that those institutions were either mis-measuring their risk exposures or, somehow, the risks they were taking were inconsistent with the risks their clients and counterparties were expecting. So the first point about dealing with crisis is to make sure we have an accurate measure of the actual risks that investments are providing to their investors, in order to properly prepare for those risks. You cannot manage what you do not measure.
Once we understand exactly what types and magnitudes of risks we're facing, it's actually not that difficult to prepare for them. One cannot legislate away hurricanes, but one can provide early warning and preparation. It turns out that if you provide localities with just a few hours notice about the impending arrival of a hurricane, you can reduce the losses from that hurricane by a tremendous amount. Insurance companies have figured this out. Similarly, if we can properly warn the various parties that they are about to get exposed to major financial dislocations, they can prepare for them, and we can reduce the degree of collateral damage significantly — pun intended.
Q: The government has ended up being the insurer in the case of the financial hurricane. How big a role do you see for the government?
I think the government has a very big role to play, but not in the way that most people have been arguing. I don't believe that government can possibly engage in the fine tuning of systemic risk. It's just not something that we have been able to do in the past, and I don't see why we'll be able to do it any better going forward. For example, the Federal Reserve System was set up for addressing bank runs and other liquidity crises. Over the last 15 or 20 years, the Federal Reserve has been very active, and yet it's been clear that they were not able to forecast, much less prevent, the current crisis.
The financial system today is far too complex for any single organization to be able to address it. One of the biggest challenges is to develop additional expertise and understanding for how the financial system imposes systemic risks that didn't exist a decade ago. I've called for the establishment of an equivalent of a National Transportation Safety Board for financial markets, sort of a Capital Markets Safety Board, whose sole function is to investigate financial "accidents" and produce reports detailing the underlying causes. The NTSB does not regulate the airline industry; all the NTSB does is examine accident sites, collect the black boxes, sift through the wreckage, and produce publicly available reports that describe what happened, how it happened, why it happened, and how we might keep it from happening again. If we did this for every single financial crisis or blow-up that occurs, then, over time, we're going to see common patterns and themes that will enable us to better prepare for these kinds of dislocations. Once we understand better what the underlying mechanisms are for these kinds of crises, we can then impose the appropriate safeguards. We certainly don't need more regulation; we just need smarter regulation.
Interview conducted and edited by Jonathan T.F. Weisberg.