This week, we learned that a hedge fund and the credit default swap (CDS) market had pushed a regional telecom company, Windstream Holdings Inc. (Nasdaq: WIN) into bankruptcy. The toxic CDS product caused havoc in 2008, and regulators have been their usual ineffectual selves in failing to grasp it since. There is no legal way to make CDS legally watertight, and Wall Street lacks the expertise to manage the product’s risk exposure. It’s time to close this casino down.

CDS represent a long-standing dream among bankers, rather as human flight did among humanity as a whole. Loans form the greatest part of most banks’ balance sheets, yet they are naturally illiquid, and their risk is impossible to manage because of that illiquidity. For decades, indeed for centuries, banks have longed to find a method whereby the risks of those loans could be offloaded when desired, so that a bank could reduce its exposure to a particular country or class of borrower, or even to an individual borrower.

When the derivatives business came along in the 1970s and 1980s, banks looked far and wide for ways to use derivatives techniques to manage loan portfolios. As I was running a derivatives business in the early 1980s, I looked at this possibility intensively myself but came to the conclusion regretfully that there was no way to create a bank-loan derivative that could be valued and managed on a sound basis.

In the same way, inventors throughout the nineteenth century devoted large amounts of money and energy to developing a heavier-than-air flying machine. (Balloons had been developed by the French Montgolfier brothers as early as 1783, but they were subject to wind and almost impossible to steer, hence did not truly solve the problem of man’s wish to fly from A to B.) In the United States towards the end of the nineteenth century, those experiments were made at the highest level, by Smithsonian Institution Secretary Professor Samuel Pierpont Langley, with government funding and backing from Assistant Secretary of the Navy Theodore Roosevelt, who during the course of the experiments became President – you can’t get higher-level backing than that!

Langley’s experiments began in 1894 and were funded by a $50,000 grant from the U.S. Army in 1898. Finally, in the autumn of 1903, the new machine the Langley Aerodrome was ready. Launched by catapult over the Potomac River, on two separate occasions, it plunged into the river, fortunately without killing its test pilot Charles Manly (Professor Langley was nearly 70 by this time, so doubtless thought discretion was the better part of valor). By future aviation standards, the Aerodrome had plenty of engine, but its control systems were rubbish.

Nine days later, at Kitty Hawk, North Carolina, the bicycle mechanics Wilbur and Orville Wright took off successfully. However, the Langley Aerodrome was not scrapped. A full decade passed, then in 1914, the pilot/aircraft manufacturer Glenn Curtiss rebuilt the Aerodrome and managed to fly it a few hundred feet. Naturally, the Smithsonian forthwith sued to invalidate the Wright brothers’ patents and put the Aerodrome on display as “the first man-carrying aeroplane capable of free flight.” Like most government projects, the Aerodrome cost a lot of money, didn’t work as planned, was lied about to the public and generated employment for lawyers. The Wright Flyer only made it to the Smithsonian in 1948, having spent World War II under a mountain in Wales, along with other British museum artifacts.

The Credit Default Swap, as invented by J.P. Morgan in 1994, is the Langley Aerodrome of the derivatives business. Unfortunately, 25 years later, we still haven’t had the Wright Brothers and so banks are trading trillions of dollars of this defective product – there were $8,346 billion outstanding at June 2018, according to the Bank for International Settlements.

Like the Aerodrome, the CDS is too unstable to fly, or in the CDS’s case, to act as a reliable hedge to a loan portfolio. It is far too easy to game CDS bankruptcy events, forcing companies that would otherwise survive into bankruptcy events and CDS payouts. That’s what appears to have happened at Windstream and has happened in several other cases. Investors who hold both debt of a company and its CDS are known as “empty creditors” and may well not act in the interests of either the company or the creditors in a stress situation. Real people’s jobs in real companies – 13,000 of them at-risk at Windstream — are being lost as a result of CDS chicanery

Another problem, and the one that prevented us from inventing a CDS product a decade before it finally appeared is that there is no solid way to calculate the appropriate bankruptcy payout on a CDS without waiting for the bankruptcy involved to be resolved in the courts, a process that generally takes years. In most CDS bankruptcy events, an auction mechanism on a small portion of the CDS has been used, but this process is artificial and takes place between a small number of creditors, each with different interests. It is hence highly “gameable,” and often produces nominal results bearing no relation to economic reality.

The uncertainty in the CDS market regarding the existence of a potential “credit event” and regarding the payout that should be made when such an event happens makes CDS the playthings of market shysters, speculators who are neither hedging risk nor truly investing, but simply speculating that their dodgy lawyer can grab more out of the situation than the opposition’s dodgy lawyer. This has greatly increased the salience of dodgy lawyers in financial markets, a wholly pernicious development for sound companies, respectable investors and the global economy in general.

There is a further problem, which is that Wall Street is incapable of managing CDS risk properly. Kevin Dowd and I discussed this problem in detail in our book “Alchemists of Loss” © Wiley, 2010. The Value-at-risk system, again pioneered by J.P. Morgan in the 1990s, works on the assumption that securities prices obey a Gaussian risk profile, which has the property that the “tails” of the risk distribution are very thin, so large losses are very unlikely. It also bases the VaR calculation on the maximum portfolio movement on 99% of trading days.

This system thus suffers from two flaws. First, most securities prices do not follow a Gaussian distribution but have risk profiles with much fatter “tails.” Second, since 100 trading days is only 5 months, losses will exceed the maximum theoretical risk of loss fairly frequently, and in conventional risk management systems, there is no way of knowing how large the losses on an anomalous day will be.

There is an additional problem, which is that some securities and derivatives positions are more un-Gaussian than others, and this differential, like everything in finance, is gameable. If a bank manages risk using a VaR system, then the bank’s “quants” will look for pathological items, which are exceptionally un-Gaussian and hence have a real risk (and potential profitability) much greater than the VaR model assumes. Essentially, traders in these pathological products can assume much more risk for each dollar of their VaR-determined risk limit – and hence potentially make much larger profits and receive bigger bonuses at the end of the year. Yes, the potential losses are larger also, but if a trader loses money, he will be fired, so he doesn’t care how much loss his ex-employer is left with.

CDS are exceptionally un-Gaussian. Their premium is generally only a percent or two, while their potential profit or loss is 100% if the underlying loan becomes worthless. Hence traders find them very attractive, and their risks in banks’ books are exceptionally understated. The solution is to manage CDS positions using a “Cauchy” distribution, a type of distribution having much longer “tails” than the Gaussian, hence appropriate for the CDS type of pathological risk. (There is another system, fuzzy logic management, that can be used for pathological instruments like Collateralized Debt Obligations, (CDOs) which have risk profile tails that are excessively fat rather than excessively long.)

Despite our book’s publication, Wall Street has failed to adopt either reform, as evidenced by the $5.8 billion reported loss suffered by J.P. Morgan in its “London Whale” disaster of 2012, which was incurred through underestimating the risk of CDS. What’s more, despite the billions of dollars spent on regulation and the 1000-page Dodd-Frank Act of 2010, regulators have also totally failed to grip the problem.

The impossibility of precisely identifying a CDS bankruptcy event and accurately assessing the payout in that event makes CDS aerodynamically unsound, the Langley Aerodrome of the financial markets. In addition, because of their exciting level of risk, CDS are exceptionally attractive for traders managed by VaR. Since financiers have less sense than aircraft mechanics, there are now $8.3 trillion of these failed experiments outstanding, with banks completely unable to manage their risk. Repeated fatal crashes are inevitable.