About Those Virgins

I don’t buy the arguments David Galland gives about the shamans and the virgins when talking about financial reform. He says:

dancing around open fires bedecked in certain animal skins or allowing shamans to sacrifice virgins were seen as effective methods for controlling the natural chaos of things.

He’s talking about policy makers in Washington of course. They are the shamans dancing around the open fires ready to sacrifice the ‘free market’ virgins in order to control the chaos that markets sometimes bring with them.

The thrust of his post – that over-regulation is poison – is something I agree with wholeheartedly. Moreover, I think David and I probably agree on a number of other things as well. One of them is that we never had free markets. In this sense I see the shaman and virgin analogy as a nice tale but a straw man argument that doesn’t really fit the bill. In my post "A more in-depth description of how elites maintain status quo ante" the virgins being sacrificed by the free-market poseur shamans dancing around the fire of crony capitalism are American citizens. What has really happened over the past quarter century is that special interests have gained so much sway over policy that even this financial reform bill crafted after a spectacular systemic failure is inadequate to deal with them.

But should we expect any different? Another point where David and I agree is where he says:

Most prevalent among the modern belief systems is that shamans of government and high finance can, by virtue of their Harvard degrees and clearly advanced intellects, effectively manage large economies. The fallacy in this notion should be evident to everyone – here in the U.S., it’s as simple as noting how everyone from the Fed chairman to almost all of the nation’s political leaders and the best and brightest on Wall Street failed to anticipate the current crisis. Any way you slice it, the lot of them were caught as flatfooted as the crew and passengers on the last voyage of the Morro Castle.

One minute, big party… the next, diving over the side of a ship to an uncertain future in mountainous, storm-tossed waves.

Clearly, the best and the brightest failed to foresee this once in a lifetime crisis. Why should we trust them to craft legislation to deal with the next crisis? We shouldn’t.

The question is what should we do then. David suggests we throw up our hands in disgust and allow the chips to fall where they may; that is the clear implication of his opening ship junket story. Give up, he says.

I take a different tack that hearkens back to my days in the technology and telecom industry. In the Internet world, any large business platform must deal with the complexity of millions of users with myriad different ways of interfacing with its service. This creates a mind-boggling complexity which can lead to catastrophic failures that result in hours of downtime or data loss and millions of lost dollars in revenue for the company and its users. We have seen these nightmares again and again. The first example that springs to mind is of AOL’s "America On Hold" fiasco when it switched to unlimited dial up Internet plans in the 1990s. We saw this type of service failure with Research in Motion’s Blackberry service going down, Google’s Gmail service going out out. EBay and Amazon have had these problems as well. And in some cases, like an infamous T-Mobile outage there was serious and permanent data loss.

Redundancy. That’s the word one learns from these scenarios. What any complex system needs is redundancy. David is replacing an all-consuming belief in the elite shaman policy maker for undying faith the free market shamans who tell us all will be well if we let the chips fall where they may. Both scenarios are not redundant. When failure strikes, it is catastrophic.

A better approach to regulatory reform is to build natural redundancy into the financial system. No system is fool-proof. But having multiple checks on system failure is better than having one.

Here’s how I would go about it:

  1. Assume an unlevel playing field: if we assume that some actors in business are naturally advantaged and will use these advantages to tilt the playing field in their favour, we must assume these actors will overreach, take on too much risk and subject us all to a catastrophic failure. Therefore, we must devise rules of fair play which are both monitored and enforced through sanctions to prevent this. This necessarily means we have to have regulations and regulators. Assuming that markets are self-regulating is folly which invites bad behaviour. Do I want law makers dictating how private enterprises should structure their compensation schemes as the Europeans want to do? No, that is what I call regulation-heavy. But, I do want consumer protections (think seat belts or child labour laws), anti-trust laws (think cases against Standard Oil or AT&T) to be robust and enforced. To me this is the first line of defence against market failure.
  2. Assume regulations are inadequate: I think the European proposals to regulate financial services compensation schemes are wrong. It’s over-regulation. And when there is bad regulation or over-regulation, one thing results: people devise (legal and illegal) schemes to get around the regulations. That’s certainly what has happened in the financial services sector. The profit motive is too large to prevent this sort of thing. And what invariably happens is that the market becomes distorted and more fragile as a result. This, in turn leads to unexpected and catastrophic failures.
  3. Assume markets fail. We should assume that markets fail – and catastrophically. That means that we often reach a critical state where panic is the order of the day and where smooth bell curve mathematical functions don’t work. The likely outcome of these situations without a lender of last resort is catastrophic business failure and economic dead weight loss. This naturally leads to political chaos, economic depression and military adventurism. Allowing markets to fail and sitting idly by is not an option. This means we must create market structures, institutions, rules and safeguards to prevent liquidity crises from becoming solvency crises. This is a primary reason the Federal Reserve exists – to act as a provider of liquidity.
  4. Assume regulators will be captured: But we should assume that these regulators will be captured by industry and fail. How else would you explain the sanctioning of increased leverage in the investment banking sector early last decade?  At one point in time, net capital rules which governed the amount of leverage investment banks could use limited them to 12:1 ratios. Policymakers changed the rule in 2004 to allow 30:1 leverage and the gross increase in balance sheets which ensued. Had Lehman been limited to 12:1 leverage, its failure would have been less problematic. How do we make this regulatory failure redundant? By creating a market structure that can handle failure. And that means, first and foremost, banks must be limited in size as a share of assets. Too big to fail is too big. Full stop.

My view is that systems fail and we need to be prepared for that failure. Throwing up your hands and acting like you can just give up and let things take their course is an invitation for anarchy and its necessary successor, despotism. What we should do is structure a system that sets basic mechanisms in place to prevent market failure but that is redundant enough so that it can deal with market failure. Moreover, one flaw need not bring down the whole system if it is redundant. We shouldn’t trust the shamans who counsel absolute faith in the infallibility of the policy illuminati. But neither should we trust the shamans who counsel absolute adherence to a non-existent free market.

  1. Traderscrucible says

    “Redundancy. That’s the word one learns from these scenarios. What any complex system needs is redundancy.”

    Exactly right on this point. We’ve spent several decades worshiping at the altar of “optimal” without recognizing that the word “optimal” has different meanings depending on the context.

    For many years we’ve operated under optimal with a meaning of highest scoring when it may have been better to think of it as least likley to break, or scores high and unlikely to break.

    But what do you mean by redundant systems in the world of finance? When you take a look at the collapse of the repo market, what does “redundant” mean? Does this mean we need the fed to be the stated lender of last resort in the repo market – much like we have the FDIC standing behind all of the bank deposits?

    Any reform measures must take into account the modern debt markets are completely and totally dependent on the repo markets for functioning. When they don’t function, we have panics. Any proposed regulations that don’t address this basic issue are nearly beside the point, because if we care about a functioning banking system, we are really talking about a functioning repo market.

    So what could make this market robust? The shock that is big enough to scare and stop this market is ‘iceberg’ risk from a large entity being of a vastly different credit rating than people expect.

  2. Tom Hickey says

    Ed, I agree with your analysis and prescriptions. I also agree that trusting the invisible hand is asking for the finger of fate. It is basically anarchism and would be worse than over-regulation.

    I would add that capitalistic systems and market economies are based on incentives. The incentives have to be correctly designed to produce efficient and effective outcomes, not only economically but also socially. Economies are for societies rather than vice versa.

    Shaping incentives is essentially different from erecting fences and gateways. The Maginot line demonstrated the folly of the latter. We need to change the incentives in order to change the results.

    “Efficiency is doing things right, and effectiveness is doing the right thing.” (attributed to Peter F. Drucker)

  3. Gbgasser says

    Well done Ed.

    I think this touches on our obsession with efficiency over efficacy. The criticisms of many regulations (probably the most effective if this is all that can be said about them) is that they reduce efficiency. The implications are that too many redundancies, to borrow your term, are costly and that we should only be concerned with how cheaply we can do something. We need to start caring only how well we do something.

    Efficiency is doing something right, efficacy is doing the right thing.

    1. Edward Harrison says

      gbgasser, I think you’re right that firms wanted to be efficient and redundancy reduces this. Individual actors will always resist redundancy until a system fails and the finger pointing begins. This is what we have witnessed. The bailouts have allowed this same focus on efficiency at the expense of safety to continue – one reason I expect another crisis.

  4. rd says

    “My view is that systems fail and we need to be prepared for that failure.”

    I am an engineer. One of the primary goals in engineering is to prevent systems from failing. As such, one of the objectives is to prevent system failure if a component fails. The redundancy concept you have at the beginning is one of the key tools in preventing a progressive system collapse due to cascading failures.

    A factor of safety is another key tool in achieving this. One of the few places in finance that really uses a factor of safety is the Graham & Dodd method of security selection. However, in finance this type of approach is generally viewed as too conservative. If engineers approached their designs with the same approach to factors of safety as the financial sector, we would all be in jail for gross criminal negligence.

    We need to make sure that the larger firms operate with higher factors of safety and that each of those firms needs to be sufficiently small so that the failure of one cannot take down the system.

    In engineering, it was found necessary over the past century to have buildings, electrical, plumbing, and fire codes in order to have safe buildings. The accounting and other regulatory standards need to be raised to the level that we currently expect of our building codes. To date, greed has prevented that. Society decided decades ago that allowing unregulated developers in building construction was unacceptable. We need to now take the same approach with the financial sector.

    1. Edward Harrison says

      rd, great reply. I will have to co-opt the part about cascading failures if you don’t mind! I think we’re on the same page about how to design systems that don’t fail. Your design feature which is made to avoid a single part failure leading to a cascade of other failures is exactly what went wrong here. I think complex systems are inherently unstable when they reach a critical state. This is what the work of Mandelbrot taught us. And that leads to this cascading. Hence my remarks regarding redundancy.

      It is this same sort of thinking behind the safety measures that should be in place in drilling by the way. And it doesn’t sound as if it was in the Deepwater Horizon accident.

      The Graham/Dodd margin of safety is also a good lens through which to think of the leverage of individual firms. I am speechless when I think of regulators permitting 30 to 1 leverage at investment banks. It was even worse at Fannie Mae and Freddie Mac. There is no margin of safety in this and you are asking for a catastrophic failure.

      1. Traderscrucible says

        In addition to Redundancy there is another idea that needs to be added – robustness. We need to be thinking about what is robust to shock and what isn’t.

        For example, imagine we stick with 30:1 leverage for investment banks. Being redundant in isn’t really a proper response – we could be redundant but what that means is that people are going to take too much risk and bring shocks to the repo market.

        We should have robust limits on banks – where if someone or some firm pushes the ratios to the legal limit, the damage can still be contained.

        Or in off balance sheet holdings, any amount of redundancy isn’t as important as simply providing robust guidelines and limits.

        This isn’t to dismiss your points, Ed, but rather to expand on them. Imagine you have a system where there is plenty of redundancy but nothing is robust – any failure could lead to a cascade of failures. Then imagine a system where things are robust but not redundant – we have a problem very similar to what we had in the last crisis – where the repo market was quite robust until a failure happened and there was absolutely no backup.

Comments are closed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More