Saturday, 30 July 2011

Two challenges for behavioural economics - one real, one fictional

It seems that some people from outside behavioural economics are, like me, getting frustrated with the lack of progress within the field.

Eric Falkenstein says here:
I read Kahneman, Tversky and Slovic's Judgement under Uncertainty in the 80's (published 1982), which mainly discussed a series of papers published in the 1970s, and found it fascinating, but now it's now 30 year old stuff and pretty boring. There's a couple hundred academically based confirmed biases which are all kinda true, but not very profound
This part of the posting is quite right. A commenter at the bottom sums up the problem with state-of-the-art behavioural "economics":
Behavioral economics is not economics but psychology. It focuses on individuals instead of exchanges and markets.
Economics makes assumptions about actors to make market predictions. BE makes predictions of actor choices just like psychologists.
The question BE must address is how do biases create market conditions. To date I have not seen BE translate into markets.
I have written something similar here; part of my ongoing research is to build good descriptive and predictive models of how markets work under "behavioural agent" assumptions. This is a good insight and I would like it to be more widely recognised.

However, Falkenstein goes on to make the following common, but fallacious, argument:
So, Sendhil likes to play to standard liberal pieties about how government can help and inequality is the result of evil and prejudice (he's big on the stereotype bias, another publication bias result gone viral). Yet he forgets the most profound behavioral bias in this literature that is delightfully recursive:

We think we are better than average at not being biased in thinking that we're better than average.
This is true especially for people who have studied this subject at college. Thus, the behavioralists neatly infer from the literature that everyone else is biased.
This is nonsense.

Experts in psychology are fully entitled to claim to understand others' biases without being biased themselves. This is not because they are magically immune from cognitive effects themselves. Of course, psychologists are fooled by pricing tricks or guided by default options just like the rest of us (maybe a little less, but not much).

No - the problem with this argument is that it's based on a blatant category error. The reason that experts are allowed to talk about biases is that while being an expert, they are operating in a completely different context than the things they are describing. We allow politicians to make laws about drunk-driving, despite the fact that many of them drink alcohol themselves: because when making laws, they are not driving (and mostly not drinking). Would you refuse to be prescribed glasses by an optician who wears contact lenses herself? No - prescribing glasses is an entirely different activity to being able to see. Could you talk in July about the possibility of eating too much at Christmas? And can a physicist write about how aeroplanes defy gravity while being personally held firmly on the ground by a gravitational force? Er, yes.

There is no contradiction in any of those situations. You can describe a process while not currently experiencing it. You can sit back and rationally consider - or model - a psychological effect, using your imperfect brain to do so, and then go out onto the street an hour later and find yourself subject to the exact same effect.

This argument is so nonsensical that - in the absence of bad faith - I can only imagine a series of intellectuals sitting back in their chairs with a whisky, so self-satisfied with cleverness in spotting the behaviouralists' hypocrisy that they smugly and immediately write it up without ever stopping to ask if the reasoning makes any sense. I am sure there's a cognitive bias to describe that, but if I claim right now to know what it is, I might be accused of suffering from it. Better wait till I'm feeling rational again.

Thursday, 28 July 2011

Notes from banking fragility talk

Here are my unedited notes from David Miles' talk on monetary policy and banking fragility tonight. This may not be of much sense or interest to everyone, so feel free to skip it if you're not into this topic.

[Here is the somewhat more coherent post I wrote before the talk]

Miles – recently wrote a paper about the benefits (and costs) of higher bank capital.

LSE is intellectual home of MPC in many ways.

Links between monetary policy and financial stability: strong and significant

Compared to pre-recession trend in 2007, this recession is almost as bad as 1929, and on current projections will be WORSE from 2012 onwards.

Changes in banking sector since previous eras – more gross debt (now being reduced), more fiscal compensation (mostly automatic stabilisers), much higher bank leverage. There would have had to be a rebalancing of some kind, but the rebalancing is now happening in the context of a financial crisis.

One challenge: policy compensations for imbalances (VAT rise, sterling depreciation) plus commodity price rises, all lead to short-term inflation. Likely to go higher still due to domestic energy prices. Squeeze on household real incomes has been very large. Some people criticise the MPC for allowing this to happen – but this is bad economics. The MPC cannot have any control over this. VAT rise, imports and commodity price increases all require real incomes to fall. In fact, if MPC had increased interest rates real incomes would have fallen even further.

Emphasises: the fall in living standards are caused by real rebalancing after the recession, not nominal phenomena.

People’s main responses to inflation: cut back spending, shop around. Few bring forward purchases or push for higher wages. Wage settlements have remained low, 2%.

Domestically generated inflation probably below 2% now.

Reinhart & Rogoff say that a 10% fall from trend growth is typical after a financial crisis.

We need to make the financial system more robust – the single best way to do this is for banks to use much more equity capital than they have to date.

Leverage has already fallen – in UK, from about 40 to 20 on average – in the US, 40 in 2007, 50 in 2008, now about 16. Spread between different banks has fallen even more.

Has suggested that equity capital should be twice as high as suggested in Basel III (relative to total assets not risk-weighted assets).

Leverage has fallen, but it’s partly because assets have fallen as well as because capital increased. Assets fallen because of lower demand for, as well as lower supply of, lending.

There’s an argument that higher bank equity will be costly (banks will reduce lending). This is not clearly borne out by recent data, but it’s a fair question: where will the equity come from?

If banks today needed to double the amount of common equity from today’s levels, you’d need a trillion dollars of extra equity. But if this is done over a number of years, it won’t pose great difficulties and won’t have substantial knock-on effects – will show why.

Liabilities of UK banks look like:
  • Deposits:
  • 14% banks
  • 5% other companies
  • 16% households
  • 18% other financial corporations
  • 8% Bonds
  • 6% other securities
  • 2% subordinated
  • 26% derivatives
  • 5% equity (of which 4% core tier 1 equity)
You could think of this change as a portfolio switch from other forms of liability.
[Is the problem that there is no demand for equity-like assets? Do people insist on liquid, fixed-income assets in debt form?]
How can we persuade people to provide equity-like financing instead of debt-like? This might be easier than we think. Ultimately, it all comes from household – though very indirectly. Investor distribution in bank shares:

  • 57% investment advisors
  • 2% mutual funds
  • 1% other
  • 3% pensions
  • 11% unclassified
  • 4% banks
  • 1% corporations
  • 18% government
  • 0% hedge fund managers
  • 0% holding companies
  • 1% individuals
  • 2% insurance companies
Similar for US and Germany.

Thought experiment: assume all bank equity and non-deposit debts of banks is owned by a single financial intermediary – an asset manager, a bit like a pension fund, unit trust, insurance company rolled into one.
Imagine banks now issued a whole lot more equity and retired a corresponding amount of bonds. The asset manager finds that in order not to change the underlying characteristics of their portfolio, they need to sell bank bonds and buy the new equity. [not demonstrated convincingly]

Claim: asset manager owns all equity but is liable for all its debts [huh?]

Like MM reasoning

Claim: asset manager owns equity and some bank debt in bond form. A switch in funding makes no difference to this manager.

Accepts that there are some problems in this: some investors only invest in equity or in debts, may not want to switch.

There is an argument (Diamond and Rajan) about why banks may need to finance a lot of their operations in demandable debt: it’s hard for people who provide debt to be sure about what the banks are doing with their money. Maybe the managers are taking high risks. One of the advantages of demandable debt is that if it looks like management are taking advantage of their control of the assets, individuals with retail deposits can respond by getting their money back. This will cause a bank run. The fear of this disciplines the management of the bank.

But most bank debt is owned by asset managers and is not demandable in this way. So this argument doesn’t really hold.

Question is: would a change to more equity funding hurt the economy significantly? Conclusion from above arguments is – not really, as long as it’s done in a gradual way. Having banks hold substantially more capital is therefore likely to make the system safer.

Notes on audience questions from twitter feed (these are in reverse order):

 Leigh Caldwell 

 Leigh Caldwell 

 Noel Bell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell 

 Leigh Caldwell