Saturday, February 25, 2012

Putting A Price Tag On Moral Hazard

I’m conscious of the fact that there are many more readers of this blog than there used to be, and not many have a background in economics.

So to introduce this subject, here’s Wikipedia’s definition of moral hazard:

Moral hazard arises because an individual or institution does not take the full consequences and responsibilities of its actions, and therefore has a tendency to act less carefully than it otherwise would, leaving another party to hold some responsibility for the consequences of those actions.

In other words, people engage in riskier behaviour since they’re not bearing the full consequences of their actions. It’s often difficult to quantify a moral hazard problem ex-ante, because you’ll only know what the cost is ex-post i.e. the cost only becomes clear once something really bad happens. and this is especially true when we’re talking about public money backing private risk taking.

Which leads us to today’s topic. From the World Bank blogs (excerpt):

Did the U.S. Taxpayer Really Make a Profit on the Bank Bailouts?

CNN Anchor Erin Burnett mocked the Occupy Wall Street protest during her show recently. Burnett asked a protester if he knew taxpayers “actually made money” on the Wall Street bailout. The protester responded that he was “unaware.”

“Yes, the bank bailout made money for the taxpayers, right now to the tune of $10 billion,” Burnett said. “These are seriously the numbers. This is the big issue? So…we solved it.”

As that exchange demonstrates, the bailout was a success or a failure depending on how you look at it. If you look at direct cash expenditures and receipts, the government has broken even or perhaps made a profit. But that doesn’t seem correct to many. Intuitively, many people from across the political spectrum have a strong suspicion that there are other costs that are not being captured by direct dollar figures.

The problem is that people don’t have a ready way to quantify what they know intuitively—that the implicit guarantee provided by the government has value, and bailouts lead to excessive risk taking and moral hazard and, ultimately, future costs for taxpayers. So instead, the focus shifts to direct measures of simple cash dollars, making the bailout appear to be a surprising success...

...In a recent paper with my colleague Joe Warburton, we investigate and try to quantify the price tag associated with the implicit guarantee provided by the government to systemically important banks. Using credit spreads on bonds issued by large U.S. financial institutions, we find that expectations of state support reduce the cost of debt for these systemically important institutions.

While there is a positive relationship between the cost of debt and risk for most financial institutions, this relationship breaks down for the largest financial institutions. Debt holders of major financial institutions have an expectation that the government will shield them from losses and, as a result, [sic] they do not internalize the full cost of the risk presented by these institutions.

This expectation of public support constitutes a subsidy to large financial institutions, lowering their funding costs…we find that the implicit subsidy given to large banks provides an annual funding cost advantage of approximately 16 basis points before the financial crisis (from 1990-2007), increasing to 88 basis points during the crisis (2008-2010), peaking at more than 100 basis points in 2008. The total value of the subsidy amounted to about $4 billion per year before the crisis, increasing to $60 billion during the crisis, topping $84 billion in 2008.

Implicit state insurance represents a significant wealth transfer from taxpayers to major financial institutions. Compared to this transfer of wealth, the accounting profit of $10 billion reported above looks rather paltry. We believe that the cost of this implicit insurance should be internalized by imposing a corrective insurance premium, thereby restoring bondholders’ incentive to monitor banks and creating a more stable and efficient financial system.

That’s a nice piece of work…though I can already imagine all sorts of problems trying to implement such an insurance scheme. Because the implicit subsidy varies with the business cycle and not against it, any calculation of premiums would understate true risk before a crisis, and probably overstate it after one.

And since no two crises are exactly alike, a moral hazard insurance fund may or may not have sufficient resources to handle any given crisis. Not to mention that any such fund would have to also invest its accumulated fund somewhere, and thus be subject to the vagaries of the markets. Even, or especially, if the fund invested in sovereigns.

There’s also the pretty pickle of deciphering what constitutes a “major financial institution”. Where do you draw the line? For practical purposes, it might be better to just implement a Tobin tax (definition here), such as is now being discussed in Europe.

One thing that the authors find, which I’ve long suspected, is that risk and financial institution size is positively related. In other words, the bigger the bank, the more likely it is to engage in risky activity. This follows on from looking at bank management efficiency, which tends to peak with medium sized institutions and tends to fall as asset size increases. Big banks are harder to manage, and harder to squeeze returns out of – which leads bank managers to pursue higher risk strategies in the name of maintaining shareholder returns.

While bigger banks might have bigger capital buffers, that capital is more at risk. That reminds me of research on SUVs and MPVs – they’re a lot safer, in that you’re far more likely to survive a crash in one than you would with a smaller car. But then they’re so big, that you’re far more likely to get into an accident in one.

Now, there are some pretty obvious applications here in Malaysia. There have been quite a few bank failures over the years, though thankfully less so now than in the past. That doesn’t mean the risk has gone however, and the buildup of debt – particularly household debt – and its increasing concentration in a few institutions that lie outside the regulatory purview of Bank Negara adds to the threat of financial fragility.

I spent much of this Friday trying to quantify how much household debt lies outside the banking system, and haven’t gotten very far yet. Even BNM is taking into account government housing loans via the Treasury, loans from insurance companies, and only two of the development institutions – leaving quite a few bodies outside the pale. Moneylenders (the legitimate ones) aren’t included, nor are any of the cooperatives or credit unions with the exception of Bank Rakyat. MBSB for instance doesn’t count towards the published total.

Yet many of these institutions have the implicit backing of the government, if only in guaranteeing their depositors through PIDM, or with public money directly. The two biggest Malaysian banking groups are government linked, through PNB and Khazanah. So are RHB (via EPF) and Affin (via LTAT). In a certain sense, the moral hazard problem here is a little less acute, as public money is already at play and so is the oversight.

To be honest, and notwithstanding all the remarks I made above about size, I’m far less concerned over these large government linked institutions than I am with those that fall outside of BNM’s direct supervision, like Bank Rakyat and MBSB, as well as the smaller cooperatives. None of these institutions are systemically important, but the basic principle is still at work.

With an implicit government guarantee on deposits, little to no outside funding (and hence little exposure to market discipline), intense competition for loans, not subject to the same rules and regulatory oversight that banks have to put up with, quite frankly they worry me on three levels – as potentially destabilising the financial system, for expanding far faster than they should, and for the role they have played in increasing household debt.

Technical Notes:

Warburton , A. Joseph & Deniz Anginer, "The End of Market Discipline? Investor Expectations of Implicit State Guarantees", SSRN Working Paper, November 2011


  1. Talking about risk, what is your opinion regarding channeling fund from EPF to Government to fund the low-cost housing loan. Is this another potential of Sub-prime mortgage loan crisis re-make?

  2. I don't have a problem with it. Both capital and return of 5.5% are guaranteed by the government, which means from EPF's point of view (and for EPF contributors), it's covered and will pay more than the current yield from MGS. 20yr MGS is currently paying only 4.1% and the last time it was as high as 5.5% was 7 years ago.

    From the borrowers point of view, it's a good deal too, as they'll only be paying the equivalent of their current rental.

    It's not sub-prime lending, because sub-prime involves paying market interest rates equivalent to the higher risk, typically double digit rates, with loan repayments to match. But that in itself raises the risks of loan default.

    That's not the case here, as loan repayments will be no more than what they are paying already in rental.

  3. I agree, however the risk here, i believe, is that the loan is given to the one who do not qualify to obtain loan from banks. The possibility of default is high and even though the (DBKL) had set up a special account (Liquidity Reserve Fund) to buy back the units from those who failed to pay loans for six months, how much can they afford to buy back.

  4. You're missing something here - given the borrowers income levels and the value of the collateral, the market interest rate charged by the banks would be too high. As well, I doubt many could come up with the necessary 10%downpayment, hence the failure to obtain loans.

    I'm reminded of the experience of microfinance, which is basically lending to dirt-poor people who no bank would ever touch. Yet repayment levels from such schemes are well over 90%, and in many cases over 98%.

    One of the biggest problems that low income households face is lack of capital and lack of assets, trapping them in a cycle of poverty. If we want to make any progress towards reducing income and wealth inequality in this country, government intervention and schemes like this are a pretty good place to start. Left to the market, nothing will happen.

  5. Few things come to mind.
    1) This seems more like major-finance considering the significant sums involved
    2) To mind, micro-finance exercises have been towards funding industriousness rather than home ownership
    3) I remember reading something about a drive for home-ownership society not having as much net positive multiplier effects as previously thought.

    Even without second guessing ulterior motives, I am cynical of how well thought out it is and most *especially the execution thereafter.

    Think PTPTN.

  6. @metalrage

    1) It's RM300 million only initially, and on a per borrower basis, it's pretty small.

    2) Quite a few microfinance institutions provide housing loans - it fills a significant gap between the haves and the have-nots, especially when viewed across generations.

    3) Didn't realise home ownership had any multiplier effects at all.

    4) The scheme's been in gestation for over two years. I think they're quite prepared to take a loss on this if necessary. As a taxpayer, I wouldn't mind one bit.

    Anybody realise that EPF can take multiple hits of this scale, and not even touch contributors' money for years?