Paul Wilmott's Blog

High-frequency Trading: Where are we and how did we get here?

Mon, 28 Jun 2010 10:30:00 --0100

"The truth is the high-frequency traders create volatility and create liquidity," said John Damgard, president of the Futures Industry Association. What he apparently meant to say was that they reduce volatility, not create it. And this was just a slip of the tongue. As Sigmund Freud observed, such slips can reveal the reality. I am concerned about High-frequency Trading (HFT) for two main reasons: Reduction of the relationship between value and price; Potential for positive feedback. Markets exist to enable businesses to raise money, to expand, to thereby employ people, and so on, for the benefit of society. This only works if the market does a decent job of revealing the true value of a company via its share price. Otherwise the market is no different from a casino, a share price may as well be given by the spin of a roulette wheel. Fundamental analysis is supposed to do a similar job. You analyze a company, study its customers, research the management, etc., and come to a conclusion. But fundamental analysis is hard work. Much easier is to run a data feed into a black box containing some algorithm, then optimize that algorithm. Your HFT black box doesn't care a hoot about the true value of a company, it only cares about what happens to the price over the next few seconds. You may spend a few months setting up this black box the first time, but thereafter you can apply it to a wide variety of markets with relatively little effort. Just re-optimize for that market. (And we know from how market players are compensated that the question of whether or not the result is long-term profitable is of second-order importance.) Not so with fundamental analysis, each market is different, each requiring the same weeks of hard work. The above wouldn't matter if the HFT boys didn't dominate the market. Is it now 70% of trades on some exchanges are HFT trades? Whenever you have a bandwagon, such as HFT now is, then you have the potential for systemic risk and feedback. Remember the last bandwagon?the credit products. How did that one turn out for the world economy? To get feedback you need a quantity of traders following similar strategies. "They all have different strategies," you say. Perhaps true for a while, but nor for long. Traders copy each other mercilessly, and since people in finance change jobs every two years it doesn?t take long for ideas to diffuse widely. But feedback can be positive or negative. Negative feedback is when an up move in a stock leads to a sell signal, and thus a fall in the price, and a down leads to a buy, and thus a rise in the price. This dampens volatility. Positive feedback is when an up begets a buy, which causes the stock to rise again, causing another buy, etc. etc. And when a fall begets a sell, causing another fall, and further selling, and? So which is it? Does HFT result in a reduction of volatility via negative feedback or an increase via positive feedback? This is an easy one. If you are a hedge fund manager which of the following would you prefer? A or B? A. Low volatility. Shares go up or go down fairly predictably. No skill is required to make money, even by the man on the street. Hedge funds can?t charge large fees. B. High volatility. Very difficult markets, experts needed and can charge large fees. If a fund does well they make a killing because of the enormous profit they have made for their clients. But they are just as likely to lose all their clients' money, in which case?nothing bad happens to the fund manager. Yes, we are in that familiar territory of moral hazard. Of course the funds want to increase volatility and they have found themselves in exactly the place they want to be to make this happen. (BTW If you want the mathematics of feedback see PWOQF2 or read the paper The feedback effect of hedging in illiquid markets, (P.Schoenbucher and P.Wilmott.) SIAM J. Appl. Math. 61 232-272 (2000). It's all about the gamma of a strategy.) How did we find ourselves in this place? Because the HFT boys cleverly played the "liquidity card" at the right time. The argument goes along these lines: "When Mom and Pop want to sell off some of their portfolio to fund their retirement then they'll get a better price if there's more liquidity. So liquidity is good." True! For the shares they've held onto for 20 years they will indeed get an extra cent. Whoohoo! Break out the champagne! So you mustn't argue with the liquidity card. The more the merrier, right? Well, no. The fact that during those 20 years their shares have lost 50% of their value thanks to the Great HFT Crash doesn't ever get mentioned. One extra cent versus a 50% fall? Hmmm. Everything in moderation. The more liquidity there is, the more you rely on its providers, and the worse the collapse when that liquidity dries up. And who is in the position to both cause this drying up, and to benefit from it? Why, it's the HFT boys again! P

Out With The New, In With The Old

Wed, 12 May 2010 10:45:00 --0100

Goodbye, New Labour, you won't be missed. And the fondest of fond farewells to Peter Mandelson, please don't keep in touch. Back to Old-style politics in which MPs and cabinet ministers have their say, no more is policy going to be dictated by a cabal of non-elected spin doctors with no understanding of the concept of 'content.' I believe that Cameron has given away more than he needed to, but the end result is a sensible compromise... Cuts to bring deficit under control: Conservative policy wins, emphasis on sooner rather than later. Good. Tax for low earners: LibDem policy to eliminate tax for those earning less than £10,000. Good. Inheritance tax: Conservatives wanted to drop this, how many times can you tax the same money? LibDems get their way, IHT tax stays. Bad. Capital Gains Tax: This could have been very, very bad. Labour eliminated the taper relief whereby tax paid depended on how long an asset had been held. But Labour never did understand how businesses work. LibDems wanted to make CGT rates more like Income Tax. That would have killed all entrepreneurial activity in the UK. Would you rather a) have a safe job, no possibility of losing money and a pension or b) risk your entire worth, risk losing your home and family and have no pension? CGT is supposed to be lower than Income Tax to compensate for the difference between being an employee and being an employer, and so encourage people to start businesses. Fortunately the LibDems did not get their way entirely. Instead CGT will rise on non-business assets such as shares and second homes. Makes a lot of sense since businesses are not discouraged. Could have been far, far worse. I still want to see the details just in case... Voting reform: LibDems want Proportional Representation, and more resulting power for them but woolly government for us. Conservatives want the status quo. Conservatives have given away a referendum on the Alternative Voting system. This is the weakest possible negotiating point they could have given away, but maybe it was a LibDem dealbreaker. We need to the see details on this. Is the resulting referendum binding in some way? Immigration: Conservatives want caps on numbers from outside the EU, LibDems want an amnesty on all illegal immigrants. This did not go down well with voters, once they started to examine LibDem policies after Clegg's game-changing TV appearance. Conservatives get their way on this one, obviously. Fixed-term parliaments: Next election scheduled for May 2015. Can't see the point in this. Mansion tax: I'll finish on this one because it has a nice quanty element to it. LibDems had wanted a 1% annual tax on the value of homes above £2m. So on a home worth £3m you'd pay £10k p.a. Fortunately this silly idea has been dropped. Why is it silly? Let's do the math on the £3m house. With interest rates so low, what is the Present Value of an annual £10k? What might happen to the value of that house? Now repeat the calculation for a house worth an arbitrary X. P

Politics, Panic and Poker

Mon, 10 May 2010 20:41:00 --0100

The Conservatives have offered the LibDems a referendum on electoral reform (albeit only on a very weak version of reform). Clearly a panic move, they didn't even wait to get a sense of people's reactions to Brown's earlier speech. I've long said that bankers ought to be forced to play poker in order to understand their own relationship with risk and return. I now think that politicians ought to play so that they can appreciate the strength of their hands and, crucially, learn to walk away from a hand to play another day. The Tories have something to learn. In classical Modern Portfolio Theory we also learn to weigh up risk and expected return. Similar principles apply in politics. Nick Clegg has to decide between low risk with the Conservatives, with a resulting majority of seats, and high risk with Labour, for who would trust them. But how high will the expected returns be? The Conservatives have offered more than they should have. I can only assume that Clegg has photos of Cameron in a compromising position. And does anyone know why this haggling is happening in public instead of behind closed doors? If this is the quality of horse trading we can expect from the current crop of British politicians then Lord help us when we have to haggle with the wider world. P

Desperate Times, Desperate Measures - How Desperate Can You Get?

Mon, 10 May 2010 17:46:00 --0100

Brown graciously offers to run the country as PM for a few more months with a LibLab pact, and then will step down after Labour has chosen a new leader. Thanks, Gordon, but do you mind awfully if we pass on this one? This is all highly disgusting and highly predictable. Brown and a few of his inner circle come up with this cunning plan, no consultation with the cabinet, just him and his favourite spin doctors. You only need to watch the 5.30pm Sky News interview with Alastair Campbell by Adam Boulton to see the Labour party imploding. Embarrassing. You expect such nonsense from banana republics. That's the bad news. There is good news, at last. What wasn't predictable about all this was that Brown would announce this decision to the public. If it was an appeal to Clegg then I think it was a big mistake. Brown is clearly flailing around, desperate to cling to power. I'm hoping that this lessens the probability of a LibDem-Labour relationship, the more natural of the two possible coalitions. Not that I'm keen on a LibDem-Conservative pairing up, I'm almost as worried by the thought of Proportional Representation as I am by the thought of higher taxes! P

Brown Out

Sun, 09 May 2010 10:04:00 --0100

Following the election result (not that there has been an actual 'result') and while waiting for the horse trading to lead somewhere, there are only two things that are absolutely certain: 1. Brown will not be PM for much longer. Whether the LibDems do a deal with the Conservatives or with Labour or whether the Conservatives go it alone, Brown will be out. This will be a condition of any LibDem-Lab deal. 2. Even as I type, Labour politicians are speaking to their agents about publishing deals. There are so many of them with knives still sticking in their backs, aching to tell their side of the story of the last 13 years. I am particularly looking forward to hearing the truth about the bullying, temper tantrums and personality disorders of the Brown years. Keep your DSM IV to hand! BTW if you've enjoyed the last few days then you'll love Proportional Representation. Under such a system every election will be like this. P

The Liberal Democrat Manifesto - Throwing the successful to the lions

Fri, 16 Apr 2010 10:28:00 --0100

On Tuesday morning I blogged about the Labour Manifesto, predicting in a Photoshopped FT front page that they would increase Capital Gains Tax (CGT) to the same rate as income tax, i.e. potentially 50%. Right policy, wrong party. On Wednesday the Lib Dems announced precisely this policy in their manifesto. I?ve no doubt that all parties will have to increase CGT to some extent in the emergency budget after the election. And an increase to match the marginal rate of 50% for the successful seems logical. But that would be a five-fold increase in CGT in an extremely short time and an almost halving of the incentive for anyone to start a business. Labour and Lib Dems policies are vindictive and completely at odds with what makes an economy function. Left-wing politicians have in their heads the formula ?Successful People = Bankers = Evil.? And in some last-days-of-the-Roman-Empire collapse have decided to throw all successful people to the lions in a vote-catching appeal to the crowds. P

The Labour Manifesto - the Deniability Quotient

Tue, 13 Apr 2010 08:16:00 --0100

I?d never bothered to read a political manifesto before today. But I suspected that some fun could be had with them. If you know that someone has a tendency to be economical with the truth then a lot of amusement can be had from figuring out how they have phrased their words to maximize the Deniability Quotient (DQ) in anticipation of accusations of lying. The DQ of the Labour Manifesto is pretty high. Take this example from Page 11 of the manifesto: ?We will not raise the basic, higher and new top rates of tax in the next parliament and we renew our pledge not to extend VAT [sales] to food, children?s clothes, books, newspapers and public transport fares.? Superficially it sounds great, doesn?t it? The casual reader interprets this as ?taxes won?t go up.? If you believe that Labour are to be trusted then you might be right in that interpretation. Ok it doesn?t exactly say that but we can trust them, no? How many advisers, lawyers and spin doctors had an input into that sentence? A more realistic assumption is that they cannot be trusted, they are political animals after all, and that sentence must be read through the eyes of a lawyer. There is enough room for weaseling around that a proper reading of this would be the exact opposite. What the sentence means is that - the percentages of income tax won?t change (i.e. the 20%, 40%, 50%) - the threshold at which they each come into effect will change, i.e. decrease - Capital Gains Tax will increase - Other taxes will also go up - The rate of VAT will increase Notice how the sentence does not specifically refer to ?income? tax. By missing out the word ?income? the effect is to suggest that all taxes are being referred to. But that can?t be the case because it?s only income tax that has ?basic, higher and new top rates? of tax. All taxes besides income tax and VAT are excluded from this sentence. This gives Labour the required deniability. I am looking forward to the Conservative manifesto!

Greed Is Good But Envy Is Bad

Sun, 20 Sep 2009 23:41:00 --0100

"Greed is good," said the Ivan Boesky-inspired Gordon Gecko in the first "Wall Street" movie. And greed certainly is one of the prime movers behind many a successful business, along with altruism, curiosity, necessity, etc. Greed is absolute. Wanting more for the sake of it. But "Envy is bad," says yours truly. Envy is relative, and all the nastier for it. It's about wanting more than your neighbour, even at a cost to yourself. There's too much envy in finance. (No kidding, Paul, what an insight!) And this has led to an escalation in salaries and, in consequence, risk taking; an unspoken conspiracy in which senior management encourage gambling by their traders so that they all get 'rewarded.' You know what though? Just divide all bankers' pay packets by ten. And since envy is relative no feelings will be hurt! I don't really do envy myself, see my Cheese blog in which I explained my retirement needs as cheese, wine, books (mostly fiction), oh yes, and a swimming pool ;-) P

Hoping For One L Of A Recovery

Tue, 01 Sep 2009 13:24:00 --0100

A few months ago I was speaking at a conference at which Nouriel Roubini gave a talk about his view on the future of the world economy. He spoke about V-shaped recovery, U shaped, W shaped, and, the horror, the horror, the L-shaped version. An L-shaped recovery is one in which there is essentially no recovery at all. Japan in the 1990s was held up as the poster child and, I hope there are no Japanese reading this, was just about the worst thing that could possibly happen to a country, to be avoided at all costs, even if it means throwing money into the system (quantitative easing) or starting another war. (That's my interpretation, not his, I hasten to add!) I gave my talk after his and, improvising, said something like this... Have you any experience of Japan in the 1990s? Well I have. And it didn't seem too bad to me. Were there hordes of people begging on the subway? Not that I recall. Was it dangerous roaming the streets for fear of being mugged? No, it seemed safe enough when I was there. Was there high unemployment and general desititution? No. More on this anon. Japan in the 1990s was, as it still is, a safe, enjoyable place, with a very high standard of living, and at the cutting edge technology wise. Not the hell hole that economists like to paint. And so I really cannot imagine what Japan would be like now if they'd had a V-shaped recovery. They'd all be communicating by telepathy and travelling via matter transporter I guess. To me the important point about the economy is not what letter of the alphabet best represents it, nor its percentage growth. After all, is it really necessary to grow at x percent per annum in order to maintain the feeling of status quo? Like the shark, which supposedly has to keep moving forward in order to stay alive. What sort of life is that? I believe what is most important is the well being of the people, and that's not the same as GDP. It is, however, closely linked to rate of employment. And that's where Japan does remarkably well. That's what governments should focus on, to L with growth! I'm a simple person. (Easy now!) I take holidays in the Isle of Man and frequent thrift shops, for god's sake! So rampant growth really doesn't do it for me. I confess that I was kind of looking forward to a world with a little bit more of something beginning with L. But it looks like we are back on the treadmill, back to the bumper bonuses. And I believe we've just missed a great opportunity for making the world a better place. (Mind you, if we get the double dip maybe we should treat that as a second chance!) P

Valuation Versus Risk Management

Wed, 01 Jul 2009 10:53:00 --0100

Valuation and risk management, two sides of the quant business, must be treated with equal sophistication, with equal respect, and with equal suspicion. And there must be closer interaction between them. In education, valuation should not be the domain of the most abstract of mathematicians with risk management its more primitive partner. In practice, quants must not produce models that risk managers cannot understand. At every stage of valuation and model development you must be asking questions about risk and robustness. It is dangerous to come up with some fancy model and only afterwards start asking questions about model error. Anyone who has ever calibrated a model knows that the methods used to mitigate model risk almost come as an afterthought, and are totally inconsistent with the original model. This need not be the case. In the CQF we treat valuation and risk management as equals. The structure of the CQF is unashamedly mathematical. Module by module we add mathematical flexibility, and in each lecture you will see model risk and model robustness discussed alongside the theory of valuation. This is how quantitative finance ought to be taught. This is the mature approach to the subject, and it will help to resolve many of the discrepancies between finance theory and finance practice. P

Policing The Police

Mon, 29 Jun 2009 08:35:00 --0100

I was at the G20 protests in London on 1st April. I took this photograph of the 'charming gentlemen' who were supposed to be keeping the peace. As we know they were instead beating people up, and slapping women around, hence their disguises. I mention this now because I've just seen a BBC story in which an MP is reported as saying that "police must modify their behaviour in an age where their actions were easily filmed by the public." Translated this means that if they are not being photographed then they can do what they like. A bit like MPs and their expenses, if no one can see what they are up to then they should be expected to get away with what they can. It seems to old-fashioned me that police, MPs, anyone with a position of responsibility in public life, should have the personality and self control to stay within the letter and spirit of the law and be role models for everyone else. This is Britain, not Italy! P

FSA: It's Worse Than I Feared

Wed, 24 Jun 2009 07:27:00 --0100

Remember my blog about Magicians and Mathematicians in which I complained about the lack of imagination in risk management? If you don't, then please take a look otherwise the rest of this blog won't make any sense to you at all! Well, I just had a very frightening experience at a conference. I used the magician example to get the audience to open up to the idea of thinking beyond the simple mathematics. I started with "What is the probability of...," and received the usual "One in 52" reply. Then the location (the magic show) was pointed out, and people changed their answer to 100%. Except that some people didn't. There were three people in the audience of maybe 100 who stuck to their original 1/52 answer and refused to budge. So far so typical. Now the frightening bit. The audience consisted almost entirely of actuaries. (That's not the frightening bit!) Except for three people from the FSA. And two of those were ones who insisted on the 'math' answer 1/52. (That's the bit that scared me!) One of them explained his reasoning. I cannot remember the details, it was quite lengthy, but the essence was that "The answer should have been one in 52 except that the magician was tricking us and so really we should ignore this factor..." (I apologise if I have got this wrong, but from the reaction of the audience I don't think I have!) Now forgive me but isn't the FSA supposed to be operating in the real world in which things are just not about pure mathematics? A world in which risk managers hide risk, moral hazard is rife and magicians do, er, magic. Isn't that sort of the entire point? If it was all about the maths then we wouldn't have the FSA, we'd use someone like the EdExcel examiners to give banks marks out of a hundred at the end of term. P

Buffett and Derivatives: Enthusiasm, Anger, Disbelief, Acceptance

Wed, 17 Jun 2009 19:37:00 --0100

Denial, anger, bargaining, depression, acceptance, according to Elizabeth Kübler-Ross the five stages of dealing with personal tragedy. I wonder if there's something similar we are going through with financial derivatives? If so, I think Warren Buffett is at the anger stage. "If you need to use a computer or a calculator to make the calculation, you shouldn't buy it," is his view on investing, famously calling derivatives "Weapons of mass destruction" a few years ago. I sympathize, and I speak as one of the mathematicians who works with derivatives for a living. In my experience there are just four stages of dealing with derivatives, and they are: naive enthusiasm, as one experiences the glorious possibilities of derivatives in one's portfolio, then righteous anger as one suffers horrendous losses, followed by confused disbelief, as one realises that no one fully understood the risk in these dastardly creations, least of all the bankers, and finally a reluctant acceptance as one admits that these things are here to stay. Let's go through these stages one by one, while I explain what each means in terms of risk. Naive enthusiasm: Derivatives are wonderful financial contracts, they allow you to finely tune your investment portfolio to benefit from your market views, assuming they turn out to be correct. My bank manager has recently been trying to sell me a contract that will give me over 5% return in one year if gold stays within a certain trading range. In market parlance this is called a double knockout quanto option. Or derivatives can be used to hedge risk from other business activities. If you regularly sell widgets to Japan you are exposed to dollar/yen exchange-rate risk. A derivative can be designed to reduce that risk for you. So far so good. Righteous anger: Who wouldn't be angry after the trillions of dollars that have been lost thanks to CDOs, MBSs, and all the other acronyms? The problem though is not the derivatives themselves, rather the way that the derivatives have swamped the market for simple stocks and shares. The notional outstanding of all derivatives globally is over a quadrillion dollars. What, you thought trillions was bad enough? You, ain't seen nothing yet! So rather than derivatives existing to help you manage risk, or profit from precise market views, the market has grown so much that derivatives seem to be there just to allow crazy leverage, risk taking on levels never seen before. And at this point the risk-management quants step in to say, don't worry we've got our fancy mathematical models that show there is actually no risk. Buffett's right-hand man, Charlie Munger, has said about higher mathematics in finance "They teach that in business schools because, well, they've got to do something." Now that really hits the nail on the head. When your competitor university across the river is charging 50, 60, 70 thousand dollars for a one-year Masters course in Financial Engineering, what are you going to do? Are you going to say you don't have any faculty that understand derivatives? Hell, no. You are going to get your smartest mathematicians together and make up a syllabus. Are your 23-year old victims, sorry I mean students, going to know any better? In 2000 I warned about the dangers of a "mathematician-led market meltdown" after seeing what had happened at LTCM and how identikit risk managers were being churned out from Masters programs, and how Groupthink was beginning to dominate risk research and derivatives valuation. I sympathize totally with Warren Buffett and Charlie Munger. Confused disbelief: I'm a great believer in education playing a bigger role in derivatives in future. But not the sort of education that we've got at the moment. I understand Warren Buffett when he says "The more symbols they could work into their writing the more they were revered." Universities are churning out many thousands of 'experts' in the analysis of derivatives but sadly they know more about the math and the symbols than they do about the markets. But again it's not the symbols themselves that are to blame, for we happily fly on airplanes designed using similar symbols, rather it's the lack of financial empathy exhibited by the multiple-PhD'd analysts, the quants, that worries me. Remember this is a mathematician writing this, but one who has been saying less is more for over a decade now. Reluctant acceptance: I've blogged in the past about the "mathematics sweet spot" for finance, where the models are not dumbed down, but equally they are not fantastically over complicated (to impress, as I expect Buffett would say). I don't think we can go back to a dark ages before derivatives and quantitative finance, but I do believe that we desperately need to rethink the type of education that those 23-year olds, soon to be in charge of your pension, are getting. Less math, fewer symbols, more commonsense, and more market know-how. And in this respect I think I'm a few stages ahead of Mr Buffett.

Avoiding Swine Flu: A Lesson From The Porn Industry

Fri, 01 May 2009 14:10:00 --0100

A couple of years ago I damaged my right hand trying to hold open the door of a London Underground train. For about 18 months I was unable to shake hands, especially with Americans. I just happened to be reading the autobiography of Ron Jeremy, legenday porn star, at the time, a book I would very highly recommend, in which he mentioned the "porn handshake." Apparently, and I emphasise that I only have his word for this, that when two porn stars meet on set instead of shaking hands, for who knows where those hands have been, they touch right elbow to right elbow. So I started doing this, because of my damaged hand, and for a while this became known as the "quant handshake." It only really caught on within a very small circle and then died out. Time to bring it back for the general population... P

Celebrity Death Test

Sun, 26 Apr 2009 18:06:00 --0100

It is almost impossible for the layperson to rationally determine the seriousness of any new disease and whether it has the potential to be the next Black Death. I'm thinking of the new Swine 'Flu, which has replaced the old Bird 'Flu as the latest Terror. It's impossible to judge because of the modern tendency to dismiss science in favour of tree hugging, hippy, holistic nonsense. And the commercial needs of newspapers who have to make mountains out of molehills to sell in this internet age. Not to mention the internet itself which encourages fear and belief in conspiracy theories. And ambulance-chasing lawyers causing us to overreact to everything to avoid lawsuits. I have no doubt that a new human virus is of more danger than global warming, and for the record I'd like to add other things that are of more pressing concern than global warming: terrorism; cyber attacks; computer viruses; (even worse) global financial meltdown; everything really thanks to the 'global village' problem. So I have my own way of determining the seriousness of any new threat to human life, it's called the Celebrity Death Test, and I hope you find it useful. The way it works is simple, if a Celebrity dies from the Threat then it is to be taken Seriously, if they don't then it's probably nothing to worry about. Bird 'Flu, fine. AIDS, not fine. I can remember when Rock Hudson died, that was the moment when AIDS became real for me. (I'm also a fan of Doris Day, read into that what you will!) You see how it works? It's just a statistics thing. If a Celeb suffers from it (and assuming it's not something that has a natural correlation with Celebrity or is self inflicted) then it is statistically significant for the rest of us. BTW your intrepid reporter is due to lecture in Mexico City in a few weeks. All being well I shall give you news from the frontline, possibly from behind a face mask. P