It is a testament to the importance of getting financial regulation right that almost ten years since the emergence of a crisis in sub-prime mortgages in the US, those countries most affected by the unfolding credit crunch are still struggling to put it behind them. The Financial Stability Board’s announcement last year about the amount of capital banks must hold against risky assets and what constitutes loss-absorbing capital (TLAC), reveals that the fundamental flaws that plagued the last regulatory approach remain (1). Financial regulation has yet to find its compass.
Too many financial supervisors consider regulation to be an exercise in ‘de-risking’ . Risk is curbed, they believe, by requiring a financial firm to put up biting amounts of capital against risks. However, risk shares much with the first law of thermodynamics – energy can neither be created nor destroyed, only transformed. When we effectively tax risk in one place it shifts to another where it is currently untaxed, like shadow banks. When we find it there and tax it again, it merely shifts once more, perhaps to non-financial institutions and so on. The logical conclusion of this process is that risk will continually shift until it settles where we can no longer see it (Nugee and Persaud 2006). That is not a good place for risks to be. The regulatory exercise should instead be about incentivising risk to flow out of dark corners and to settle where it can be best absorbed (Persaud 2015).
Another fundamental flaw with the modern approach to risk management is the notion of risk-sensitivity on which the new TLAC requirements are built (Persaud 2009). Banks don’t topple over from doing things they know are risky. They topple from doing things they were convinced were safe before they turned risky (Schularick and Taylor 2009). Against loans they deem risky, banks demand extra guarantees, collateral, interest, and repayment reserves. Against their reported risk-weighted assets they were never as well capitalised as just before the crisis. It’s not the things you know are dangerous that kill you. Under the risk-sensitive approach, they had (and will have) the least capital against those assets that they thought were safe just before they turned bad. Banks appeared undercapitalised after the crisis but not before. Risk sensitivity falls into the post hoc ergo propter hoc trap.
This flaw of risk sensitivity is even worse than it appears. In the regulatory-approved practice of risk-sensitivity, banks, corralled into using the same risk models and data sets, end up buying the same assets that the models estimate had the best yield-to-safety ratio in the past. These risk models then force them to sell these assets at the same time when there is some disturbance in market volatilities and correlations (Persaud 2000). What has been generously called the Persaud Paradox of market-sensitive risk management – the observation of safety creates risks – leads to two related but separate problems (Davis 2005). It made the safe assets risky and it created a fallacy of composition that regulators overlooked: the process of trying to rid individual institutions of risk created systemic risk because they were selling the same assets at the same time.
Another fundamental error is the treatment of risks as if different risks can be added up together, and the aggregate amount of risk, whatever it is made of, can be hedged with capital. Different types of risk require different hedges. While capital is appropriate for credit risk, it is not the best hedge for all other risks. For instance, the way to hedge liquidity risk – that is, the risk that, were you forced to sell an asset tomorrow, would fetch a far lower price than if you could wait to find an interested buyer – is by having long-term funding to tide you over. If markets became illiquid and you were short-term funded, no tolerable amount of capital would save you (Goodhart and Perotti 2015). Credit risks, on the other hand – the risk that someone defaults on payments to you – rise the more time you have. Matching credit risk to long-term funding would not hedge credit risks. The way to hedge credit risks is to diversify across uncorrelated assets including loss-absorbing capital.
In Reinventing Financial Regulation I present a resolution to these fundamental flaws.
- The key mechanism of any solution has to be a change of incentives.
Although many see bad and unethical behaviour in the crisis, most of the behaviour that contributed to the crisis was incentivised. It would have taken place anyway because of these incentives. Banks sold credit risks to institutions that had no natural capacity to hedge or absorb credit risks. They did so because they had to put up capital against credit risks, and the special purpose vehicles, insurance firms and hedge funds that bought them did not.
In place of the credit risks that banks could have diversified across, their customers, the banks, bought illiquid instruments that they had no long-term funding to back when markets froze, like long-term mortgages and loans to private equity investments. They did so because illiquid assets offered higher yields (the liquidity premium) but capital requirements were driven off separate credit ratings. The banks were incentivised to create instruments that would attract AAA credit ratings, however indecipherable their make-up and however illiquid they could become. Locking up bankers is a satisfying rallying call but the real challenge is to change the incentives that create the booms that lead to the busts. (2)
- Financial institutions should be required to put up capital or reserves against the mismatch between each type of risk they hold and their natural capacity to hold that type of risk.
Risk capacity is not risk appetite. It is not determined by your ability to measure the economic cycle, which we have collectively proven bad at. It is the ability to naturally hedge a risk. Risk capacity takes a structural approach to risk management rather than a statistical approach. Let me illustrate risk capacity with the following example. Part of the extra return from equity market investments, over that available on cash, is compensation for daily price volatility. Institutions with long-term funding or liabilities, like life insurers and pension funds have a natural capacity to earn this risk premium because unlike a bank or casualty insurer, their liabilities do not require constant liquidity and stability before they become due.
If capital was set against mismatches between risk capacity and risk taking, most of the capital that life insurers and pension funds have to put up would be for concentrations of credit risk, not liquidity risk. A bank on the other hand, with many different customers, has a natural capacity to diversify concentrations of credit risk. Until they revert to their original business model, most of the capital they put up would be against liquidity mismatches (Calomiris and Watson 2014). Banks would be incentivised to sell good-quality credit but low-liquidity assets, like infrastructure bonds, to life insurers. They would also buy low-quality credit risks, like corporate bonds as they are better placed to hedge these.
The direction of these risk transfers would strengthen the financial system. They are in the opposite direction to the transfers witnessed in the run up to the financial crisis when risks ended up where there was greatest ignorance of them and least capacity to absorb them. Systemic resilience would not come at the cost of more capital because resilience is coming from repositioning risks to where there is a natural absorptive capacity and therefore where, if these risks blew up, they would not bring down the entire financial system. It is time for risk-sensitivity to move over in favour of risk capacity.
The TLAC announcement confirms that this is not the route regulators are following. Indeed, under Solvency II, regulations due to come into force on 1 January 2016, the natural buyers of long-term investment instruments – life insurers and pension funds – are being forced to shun them because these instruments are not liquid or stable enough in the short-term, even though life insurers and pension funds do not need short-term stability and liquidity (Persaud 2015a). Perhaps if long-term investors were able to fund more long-term investment we may be better able to tackle structural stagnation. The reinvention of financial regulation I propose would treat savers better, make the financial system more resilient, and raise economic growth.
Brunnermeier, M K, A Crockett, C A E Goodhart, A D Persaud and H S Shin (2009) “The fundamental principles of financial regulation”, Geneva Report 11, ICMB and CEPR.
Calomiris, C and M Watson (2014) “Corporate governance and risk management at unprotected banks: National banks in the 1890s”, NBER Discussion Paper, No. 19806.
Davis, P (2005) “Risk management is a risky business”, Financial Times, June 20.
Goodhart, C A E and E Perotti (2015) “Maturity mismatch stretching: Banking has taken a wrong turn”, CEPR Policy Insight 81, May.
Nugee, J and A D Persaud (2006) “Redesigning regulation of pensions and other financial products”, Oxford Review of Economic Policy, 22(1): 66-77.
Persaud, A D (2009) “How risk sensitivity led to the greatest financial crisis of modern times”, VoxEU.org column, 7 October.
Persaud, A D (2000) “Sending the herd off the cliff edge: The disturbing interaction between herding and market sensitive risk management practices”, Jacques de Larosiere Award, Institute for International Finance.
Persaud, A D (2015) Reinventing financial regulation: A blueprint for overcoming systemic risks, in press.
Persaud, A D (2015a) “How not to regulate insurance markets: The risks and dangers of Solvency II”, Peterson Institute for International Economics, Policy Brief 15-5, April.
Rose (2015) “The risks of de-risking: Conflicting pressures on financial institutions”, Banking and Finance Disputes Review, May.
Schularick, M and A M Taylor (2009) “Credit booms gone bust: Monetary policy, leverage cycles and financial crises, 1870-2008”, NBER Discussion Paper, No. 15512.
(1) The Financial Stability Board, created by the G20 nations, announced on 9 November 2015, that the most systemically important lenders must have a total loss-absorbing capacity, including bail-in securities, equivalent to at least 16% of risk-weighted assets in 2019, rising to 18% in 2022.