Dancing Around the Risk Question

Reflecting on my last two posts – price caps, loan structure requirements, underwriting rules – discussing any of these puts the cart before the horse. We know we want to rein in risk without cutting off access to credit that is not too risky. But how much risk is too much risk when it comes to credit? 

I began posing this question to audiences at one of the very first talks I gave as an academic (a 2005 talk about predatory mortgage lending), but while most of my talks generate plenty of responses, not once has a single audience member attempted to answer this question.  

It is a remarkably difficult question to answer, one that varies with the expected costs and expected benefits, to borrowers, lenders, and society, of each extension of credit. Moreover, actual future costs and benefits are often unknown and perhaps unknowable (meaning we are dealing with uncertainty, not merely outcomes with known risk distributions) and incommensurable (meaning tradeoffs are difficult).

A home purchase mortgage that allows a family with children to move to a neighborhood with better schools may provide lots of benefits, but the costs that would be incurred if the family were to lose the home are also great. (This is not to say that homeownership opens access to better schools; at least one study has found that low-income families who buy homes often do so at the price of moving to a neighborhood with lower-quality services). A cash-out refinance mortgage that allows a homeowner to buy an SUV may provide fewer benefits than one that allows a homeowner to start up a new business, but perhaps not if the new business has little potential to flourish. 

Such variation in expected costs and benefits might seem to make the appropriate degree of risk a question that borrowers and lenders should decide for themselves. But the recent crisis demonstrates that leaving the decision up to borrowers and lenders produces too much risk, at least in the mortgage market.

We also claim that apart from the expected costs and benefits of any particular risk-laden decision, people have heterogeneous “preferences” for risk that we should honor by allowing them to make their “own” risk choices. Since the crisis, there has been some recognition that borrower and lender risk choices ought to yield to externalities, such as those wrought by foreclosure. But there are two deeper issues here.

The first is the degree to which people do not really make their own financial risk choices. Although the recent boom and crash was due in part to investors intentionally engaging in speculation, many owner-occupiers who “chose” the riskiest mortgages were demographically those we know are the most financially risk averse. As I have detailed elsewhere, people have great difficulty matching their risk “preferences” to their mortgage choices. The same is true for investment decisions. For example, people who invest disproportionately in their employers' stock are rarely satisfying a taste for risk; they often do not realize that they are making a risky choice. When financial risk choices are mistakes rather than instantiations of preferences, there is no reason to respect these choices. 

The even deeper issue is the normative dubiousness of respecting individual risk preferences even where people can operationalize them well. Risk-seeking is strongly correlated with impatience, and we typically view impatience not as a "preference" to be respected but a character flaw that should be altered or a brain malfunction that should be fixed. As I explained in an earlier post, executive function, including self-control, has strong effects on individual financial outcomes.  Exposure to alcohol in utero and to highly stressful environments early in life (abuse, household instability such as through multiple foster placesments, etc.) can damage the development of executive function. We view poor executive function and low levels of self-control as problems, not preferences.  

The claim that we should respect heterogeneous financial risk “preferences” thus seems little more than an excuse not to engage in difficult societal tradeoffs. The truth is that we do not know what the right amount of risk is.   

Yet enacting policies to rein in risk without knowing what level of risk we want is stumbling around in the dark as surely as is engaging in financial education or creating disclosures without knowing what financial behavior we seek (discussed in my prior posts). This is not to say that we should wait until we have answered the risk question before taking action; the perfect should not be the enemy of the good. But we ought at least to be trying to answer it.   

How much risk is too much risk? I don’t have the answer to this question, but I invite you, fine readers, to try.

Comments

[Creditplumber]

You raise some good points, Lauren.

Societies have collectively bought into the fallacy sold by 25 years of cheap banking - and which governments bought into hook,line and sinker - that banks can "mitigate" risk by moving risk from a borrower or a "promise" in the form of a financial guarantee. Cyprus this weekend has painfully proved that theory wrong to small bank deposit holders.

For entities public or private, the answer is relatively simple - establish a risk framework with clear risk limits delegated to individuals based on a level of risk an entity can afford to manage, measure and maintain. Tie these risk limits to incentives and remuneration to ensure people don't just take jobs because of a) charisma 2) past performance 3) who they know 4) skilful at obfuscation/opacity (politics). This is Enterprise Risk Management, pure and simple. But it has been missing from most entities who believe growth was driven by debt, risk-taking and a singular view of the world created by senior execs (or the "powerful" -see banks, resource extraction companies and democratically-elected governments.)

It is absurd to think the consequences of risk-taking can be "transferred". Yet that is where "developed" economies appear to be.
The establishment of risk appetites (and the ability to fund these contingent liabilities should they occur) is something "easy credit" has allowed households and entities to ignore. The future will need to be different.

[David Shemano]

Lauren --

You do raise good points, but you skip over the "meta" problem. If individual choice can lead to bad decision-making, which leads to the conclusion that a political solution is necessary to limit individual choice/consent, you are necessarily assuming that the very individuals who cannot make good decisions on matters that affect them directly will make good decisions when they enter the voting booth. Do you believe people make better decisions in the voting booth than in everyday economic matters?

[Creditplumber]

Lauren,

Big questions! No, I don't think people make better decisions in the voting booth than in everyday life. There is often a compounding of heuristics and biases relating to social feedback (parents, friends, employment, etc) that influence voting more than in everyday decision-making. However, voting is an infrequent event. Daily decision-making isn't. Reading Kahneman, Tversky,Thaler & Sunstein helps us to understand a little of the psychology of decision-making in todays world. One point Kahneman makes well in his latest book is the more you raise peoples remuneration beyond the median, the more they believe successful outcomes come from their skills in decision-making, rather than luck or serendipity.
Over the course of the last 30 years we have increasingly seen pay (bonus privatised, malus socialised) become further removed from long term performance. This has been assisted by the fallacy, germinated and spread by investment and commercial banks, that losses from financial risks could be "100% removed" from an entity (or person in respect of 100% mortgages) without recourse by financial engineering. This is plainly absurd. It's like suggesting people can go much faster in cars with babies on board because of considerable child seat & car safety science over the last 50 years ago.
Speaking from the perspective of corporate entities, this therefore requires a review of how risk is identified and measured and the benefits (feedback mechanisms with counterparties/ cost of capital) of externalising a proportion of identified costs associated with these risks. I was an underwriter at AIG when BP decided to self-insure completely for reasons that included cost of capital. I suggest they lost the feedback mechanism that may have informed them of root causes of engineering failures associated with other energy losses worldwide which ultimately influenced losses in Texas and Gulf of Mexico. Following the Gulf of Mexico loss, executive remuneration was amended to include environmental KPI's alongside financial KPI's.
To wrap up, "how much risk is too much risk?" is a great question that should be a central discussion point in all respectable boardrooms. The ability to manage risks and their outcomes effectively should directly affect the culpability of executives in the event financial performance is undermined by contingent events whether endogenous or exogenous or within or outside the control of the Board.

[Knute Rife]

You can't set a level of acceptable risk if you don't have well-articulated goals, and our societal goal statement currently consists of little more than, "We want everything to be nice." We want people to get more education, and so we subsidize the risks involved, but we never bother to define what we want people educated in or why. Do we really need more English, psychology, and law graduates? We've needed more machinists, tool and die operators, and pattern makers for decades, but our subsidies aren't producing them.

We have the same problem with housing. We want people to own homes, but why? Not just lower income folks but middle income as well can rent into far better neighborhoods and schools than they can buy into. And is the average suburban neighborhood any more socially cohesive than the average apartment complex? How useful is to a family to buy a house if the CC&Rs forbid everything meaningful, from individual decoration to vegetable gardens (Don't EVEN get me started on what a bad idea our 60+ years of driving people off the land has been. Severing a population from a food supply is a recipe for disaster. If you ask where kids where food comes from, and all they can answer is, "The supermarket," ....)? Once again we're just throwing money at a hortatory statement. We have no real goals, so we have no way of determining what reward is worth what risk.