The Economist on Adverse Selection and Moral Hazard

The Economist is running a series on classic articles that have transformed economics, starting with George Akerlof’s 1970 “Market for Lemons” paper. Akerlof catalyzed the field of information economics by pointing out possible consequences of asymmetric information in the case where one party to a transaction has more complete information about product quality than the other before the transaction takes place. In this case, symbolized by the market for used cars, buyers will be skeptical about seller assertions of product quality, fewer used cars will be sold, the cars sold will be of lower quality than in the absence of the information asymmetry, and there will be less welfare created than otherwise.

This adverse selection problem can be costly, and it’s a pervasive epistemic characteristic of life. Akerlof’s model and analysis brought those costs into brilliant focus. They also planted seeds of new ideas and new research: Michael Spence on signalling, Joseph Stiglitz and Michael Rothschild on mechanisms for screening that induces people with different traits to make different choices (and thus achieves what’s called a separating equilibrium instead of a pooling equilibrium):

Suppose a car insurer faces two different types of customer, high-risk and low-risk. They cannot tell these groups apart; only the customer knows whether he is a safe driver. Messrs Rothschild and Stiglitz showed that, in a competitive market, insurers cannot profitably offer the same deal to both groups. If they did, the premiums of safe drivers would subsidise payouts to reckless ones. A rival could offer a deal with slightly lower premiums, and slightly less coverage, which would peel away only safe drivers because risky ones prefer to stay fully insured. The firm, left only with bad risks, would make a loss. …

The car insurer must offer two deals, making sure that each attracts only the customers it is designed for. The trick is to offer one pricey full-insurance deal, and an alternative cheap option with a sizeable deductible. Risky drivers will balk at the deductible, knowing that there is a good chance they will end up paying it when they claim. They will fork out for expensive coverage instead. Safe drivers will tolerate the high deductible and pay a lower price for what coverage they do get.

Notice the intersection of this idea with price discrimination; the information economics angle on it highlights the extent to which one party knows something germane to the transaction that the other party does not.

A logical next step in the analysis, then, is the mechanism design question, which I think of as an institutional question (although most mechanism design theorists probably don’t): what rules should structure the transaction to minimize the negative impacts of the information asymmetry? This is where those differences in insurance deductibles come in (and this is similar to the reason why a price discriminating monopolist creates more consumer surplus and producer surplus than a single-price monopolist).

This line of research also led to the development and formal analysis of the idea that there are two broad categories of different types of asymmetric information: adverse selection, which is asymmetric information that affects decisions before the transaction, and moral hazard, which is asymmetric information that affects decisions after the transaction (ex post). Again insurance provides a good example, but for a different reason. Knowing you have car insurance may induce you, at the margin, to drive more quickly, increasing your fun but also increasing the risk of a crash. The insurer can’t observe your ex post behavior, but would like to structure the transaction so you don’t take those risks that would cost the insurer more money. Here the deductible also provides a mechanism to reduce your incentive to drive quickly, because if you do you bear more of the cost. In 1975 Sam Peltzman wrote about the unexpected negative consequences of devices such as seat belts and athletic helmets as examples of moral hazard, and Gordon Tullock famously said that to negate moral hazard all steering wheels should come equipped with a big spike facing the driver.

The most salient recent application of information theory and moral hazard has been to analyzing the sub-prime mortgage derivatives, the creation and trading and insuring of those derivatives by financial institutions and insurance companies, and the resulting financial crisis of 2008.

Of course it’s overly simplistic to label asymmetric information a “market failure” and advocate regulation without exploring the institutional alternatives that people come up with to reduce its costs. Insurance deductibles are an example, as are warranties on used cars. More simply, individuals can keep a detailed record of car maintenance receipts (itself a costly activity, but can be cheaper than the reduction in sales price when you want to sell the car). New businesses also emerge to reduce information asymmetry — this validation of vehicle quality is the core of Carmax‘s business. Digital technology has also created myriad reputation-based mechanisms for mitigating information asymmetries, which is one way of thinking about how the “sharing economy” creates value that didn’t exist before.

3 thoughts on “The Economist on Adverse Selection and Moral Hazard

  1. Information asymmetry, in my view, is the explanation for why high-frequency trading is degrading, rather than improving, the efficiency of financial markets. At the shortest time scales, information is _always_ asymmetrically distributed. Recent innovations (e.g., temporal buffering) are making progress in mitigating this problem. With the SEC’s approval of the new IEX stock exchange, I’m hoping to see more progress on this front.

  2. I don’t think that this analysis is complete enough. Large informed traders are the bane of liquidity providers such as HFT firms. The HF part is largely a countermeasure to getting screwed all the time by the big investment banks.

  3. I have no wish to take sides in the HFT arms race. Rather than decide who is at fault, I think it is more productive to think about how to make the markets more efficient. When orders and information are flowing at the same speed, racing is inevitable. But adding a speed bump allows information to flow at whatever speed technology will allow, while trading takes place at a slightly slower pace. The result is that trading takes place in an environment (i.e., a time-scale) where public information is more widely distributed, allowing price discovery to operate more efficiently.

    Since this is Lynne’s blog, I will mention two analogies from bicycle racing that I have used to try to illuminate this problem. One is the use of sponges — soft speed bumps — to keep racers from cutting the corners on a bicycle track. The sponges are safe and self-enforcing, and provide the right incentives — including what lawyers might call efficient breach. They are a much better technology than a trading (racing) halt, or “circuit breaker.”

    The second analogy is the contrast between a sprint on bicycle track (look for one on youtube), and a pursuit. You would think that a one-on-one sprint would be fast, but there is a Zugzwang problem: a first-mover disadvantage. Sprinters go slow or even stop completely (there is actually a rule against bicycling backwards), in order to avoid giving an advantage to their opponent. In a road race, the first-mover disadvantage is primarily due to the wind, but on a track much of the disadvantage is informational: the second rider is better able to monitor her opponent. By putting riders on opposite sides of the track, the pursuit restores informational symmetry. Financial markets suffer a similar problem with liquidity providers, and maker/taker compensation schemes are a symptom. I think the biggest advantage of Eric Budish’s proposal for frequent batch auctions is that call auctions (like the pursuit) restore the symmetry between makers and takers, and thereby avoid the liquidity lock-ups that we sometimes see today. In a call auction, we don’t need to think about who goes first.

    Sorry; this is a big topic to explore. But I believe Akerlof’s insights into information asymmetry lie at the core of the HFT problem. And financial market regulators could learn a lot from the organizers of bicycle races.

Comments are closed.