An Omen – the start of trading on LIFFE David Harding was in at the beginning of the futures markets in Europe. He was a graduate trainee at U.K. institutional brokerage Wood Mackenzie in the gilts department, and so it was natural that he should be despatched to the floor of London International Financial Futures Exchange (LIFFE) as it opened in the Royal Exchange opposite the Bank of England in 1982. After a year there he had learned to draw his own charts, and had devoured books on technical analysis, as a bright physics graduate from Cambridge University would. He was looking to apply some level of scientific analysis to chartism. After a while he came to the conclusion that he needed to be in commodity futures because that way he could be in any market when it became active, rather than being maroonedin one market when it became dull to trade. "I was quite young and quite ambitious, so when I got offered a job at Johnson Matthey Commodity Brokers I took it," says Harding.
He was given an entrepreneurial position to ring stockbrokers to sell the benefits of using stock index futures. After 3 months of tele-sales effort to "every broker in Britain" he had exactly one client. As Harding had successfully sold evening newspapers by 'phone when he was at college to supplement his income, he concluded that it probably wasn't his sales technique that was at fault. He persisted and was still there after a year when on a marketing trip to Edinburgh he got a telegram at his hotel. Johnson Matthey Bankers, the parent company of his employers, had been nationalised. As this was in 1984 in Thatcher's Britain it was quite a shock. There was a banking scandal and the bank had got into serious trouble, and as one of the five members of the London gold price fixing it was deemed too important to be allowed to collapse.
Around this time Harding was being solicited by Sabre Fund Management to come and join them in their Commodity Trading Advisor (CTA) business. "I went to Sabre because I didn't want to sit in an investment bank and make money. I wanted to know if you could do it from outside the markets looking in. "Could you be on a desert island and make money trading?" was the question I was asking myself." Partly this was Harding the scientist exerting an influence, as the efficient market hypothesis was a hot topic amongst academics that studied markets. "This was a time when the application of mathematics to markets was coming to the fore – there was stochastic calculus and option valuation, as well as the glamour of efficient market theory itself," says Harding. He got on top of option pricing and recognised the appeal to mathematicians of looking at markets exhibiting Brownian motion. "It was satisfactorily difficult. It was like a prize sudoku puzzle, with the added appeal that you got paid a load of money for looking at it, " he chuckles professorially.
"With the EMH under consideration I was very interested to know whether the very antithesis of that, technical analysis, had any truth in it. So I went to Sabre very much in the spirit of intellectual curiosity." According to Harding the prevailing thinking at places like Wood Mackenzie was that there was no intellectual respect for technical analysis. There were a hundred analysts there, consisting of various types of serious people, and the technical analyst was isolated at the end of a row in an out-of-the-way part of the office. His craft was considered intellectually light-weight by his peers and the press. " I remember being visited by the FT's Barry Riley in 1988. Within a week there was a side-swipe in print along the lines of "that's more than the whizzkids with computers can manage," " recalls a somewhat wearisome Harding. "But the thing is he and all the other serious investment professionals nodding sagely in agreement were wrong. No amount of looking knowledgeable and smug will make you right if you are wrong in the first place."
This attitude was something that Harding was to get used to over his career. "For eighteen of my 25 years in markets I have received intellectual scorn and derision from everyone – business school professors, senior investment professionals to experienced journalists. But thankfully now it's different: the battle is won. The enemy is routed. All we are doing now is mopping up the stragglers."
David Harding spent two years at Sabre Fund Management, each day drawing hundreds of charts by hand, like a true craftsman. Every chart was bound into big leather folders, and in turn each chart pattern was copied into other folders. He likens it to an old-fashioned publishing house. Harding noted without irony that the company was run by accountants, andthat "there certainly was method in what was done there." He continues "I certainly regard my time there as the foundation stone of my credentials as an empiricist. There is nothing like drawing thousands of charts by hand to fix them in your mind. In fact I regard this phase of exhausting taxonomy of technical analysis as being like the relationship used to be between biology and taxonomy in the life sciences. Until something like 1830 you had gentlemen scientists collecting leaves and putting them into folders, and it wasn't until Darwin that he and others started putting some order on it. Only by arranging data and putting it in order can you get any pattern out of it. What AHL* and others have done is go beyond the taxonomy and turn trading into a real science."
At Sabre there was only one trader, and he is Harding's only co-executive director at Winton, Martin Hunt. Hunt later went on to head up AHL's trading desk. So Hunt and Harding have worked together for thirteen of the last nineteen years. At Winton Capital Management Martin Hunt is Chief Operating Officer (and Harding is Chief Executive Officer). Winton has a non-executive director, and minority shareholder, Osman Murgian. Murgian had also been an early stage investor in, and director of, AHL.
From his time at Johnson Matthey Commodities Harding knew the only other CTA based in London, run from Brockham Securities, the sugar-broking company of the Adam family. The Adam family wealth was derived from bringing the sugar crop of Mauritius to London, so the businesses included ship broking and sugar merchanting as well as commodity broking. The bright son of the family member running the commodities joined the firm – his name was Michael Adam. According to David Harding the younger Adam was brilliant and highly-strung, and the two had much in common – Adam had just come down from Oxford with a physics degree, and Harding had graduated from Cambridge with an Honours degree in Natural Sciences specialising in Theoretical Physics.
At that time Brockham Securities ran managed accounts across a range of markets – coffee, cocoa, sugar, aluminium, zinc and copper – based on a simple technical decision rule related to five consecutive closing higher highs. Being analytical and technical Mike Adam wondered whether varying the number of highs would make a difference. So he wrote a computer program that could back-test that and other questions that arose, like "how far through the high is a significant amount?" "Today we know those questions as determining the parameters," says Harding, "but there was no one else doing it in London then." The output from some of the routines could be displayed as a heatmap. "I was very impressed with what Mike was doing, as were others. They were rather tucked away in Wimbledon [in suburban SW London]. I'd done three jobs by then and learned quite a lot, so wearing my salesman's hat I knew we could sell this."
It was clear to the younger generation that there were synergies between what Sabre were doing in technical analysis, and the analytical capabilities of Mike Adam's system. "In a moment of inspiration Mike Adam said to me "why don't you come over here and join us; my father is close to retirement and you could become an equal partner in the firm with me and Marty [Lueck]." " says Harding. Having left Sabre, it was unfortunate for Harding and the younger Adam that Brockham Securities did not last too much longer. The trading system failed for the first time and the trading accounts lost a lot of equity. Being a man of honour the older Adam refunded the account holders with equity from his own firm, but as a consequence the firm folded.
The neophyte systematic traders, Harding and Adam persuaded Martin Lueck, working in a research capacity at Brockham to join them in a new venture to implement their trading system independently. The new venture, *Adam, Harding and Lueck wasformed in 1987, and through time the three founders settled into their own roles. David Harding was in charge of research and Mike Adam and Marty Lueck built a technological capability that was second to none in the City. Of course in those times of proprietary computer languages and standalone box solutions from hardware manufacturers it took major investment to stay on top of the curve in computer systems. Those were big spending decisions to make. Laughing at his 26-year old self, Harding says that they were callow enough to declare at the time that they should have the best technology for its own sake. Not a sentiment he shares now.
Over the period 1991 to 1994 the three founders sold the firm in stages to ED&F Man, and the AHL Fund has been a cornerstone of the success of Man Group ever since. Mike Adam eventually left to pursue his interests in technology, though not before some "strategic differences of opinion" had arisen between Harding and the other two principles of AHL. Harding now says that the others wanted to concentrate on the technology, while he ran the research side. Indeed Harding headed a unit called Man Quantitative Research that was separate from AHL.
Harding's view was that the core of the business was money management, and that his passionate interest in the intellectual side of the business should be brought to bear in building the firm in a charismatic fashion, whereas some of his fellow directors at AHL wished to pursue a more institutional and orthodox approach "I absolutely believe that there is a role for structure and process, or "form" if you will, in the development of successful money management businesses, but form should not take precedence over content. There did seem to be some danger that management by PowerPoint would dominate."
David Harding recognised that the AHL partnership was no longer as fruitful as it once had been, and in addition, he was unable to push on at Man as he would have liked. He would have to find a way to set his own business agendas. So David Harding left Man to set up Winton in 1997.
Just after he set up as Winton in 1997 Harding describes going to a conference. "I met this potential investor, and I said to him that we were an exchange-traded directional global macro fund, he said "I don't invest in CTAs," and walked away." Harding says that it is not so much how you position yourselves within the universe of funds: whatever you do, investors will simply put you back in the box they label "CTA".
Harding continues the story of his successful CTA business: "We were gathering assets quite nicely until 1999, then for a while it seemed no one wanted a trend-following CTA. We didn't raise any money at all in 1999 because everyone said they didn't want another me-too trend follower. They said they have that strategy covered through other managers, whether it was Chesapeake, or John W. Henry or Rotella. Their portfolio construction experts say they needed something which doesn't correlate with trend-following. So they put their money into Niederhoffer who could tell a story about being counter trend-following, and other sundry people who since disappeared completely," explains Harding.
David Harding has another explanation for his experience of capital raising at that time. "The other thing was that most of the people I talked to didn't know what I was talking about. It was a hard sell in 1999, the potential investors just didn't know enough. I had failed to sell the idea of a full-blown trading investment management business (like Tudor, or Renaissance) to Man in 1995 and even by 2002 I was still failing to sell the idea to investors."
His experience was little better in the years after. "There is an English idea of succeeding effortlessly, but that has not been my experience," discloses Harding. "I have been trying hard all along to sell what I have been doing. In 2000 and2001 I got a lot of media coverage to get my message out there. But still in 2001 and 2002 a lot of investors chose other systematic CTAs over "hauteur" Winton Capital, and these decisions are made on a logical basis aren't they? I don't think I did anything wrong, but my timing wasn't right in trying to sell then," concluded the Winton main man.
Harding has also identified another, possibly serious, issue that inhibits even dedicated investors from gathering capital. "The more senior people, that is the more experienced hedge fund investors, can discriminate on the basis of what you say [as opposed to other means of analysis]. However unless you are reasonably powerful you don't get to meet the senior people at the funds of funds or wealth management houses. I see this as a significant limitation of the way some investors operate," concludes Harding.
The track record of Winton Capital Management is impressive. Compound returns since inception are 21.47% per annum with an annualized standard deviation of 21.90%. In addition monthly returns show no correlation with the S&P500 (correlation -0.04), a large positive annualised alpha (24.62%), and an information ratio of 0.65 according to IASG data.
David Harding has every reason to be proud: "if you look at MAR, over 1-year, 2-years, 3-years, 5-years and 7-years Winton is the top performing CTA in the world. There is no doubt about it," he states. The evidence backs him up – see table below. In football terms that puts Winton Capital at the top of Division One according to the owner/founder. The Premier League is beyond that and is composed partly of firms that were the top traders when David Harding started in the business: in 1987 the top trader was Paul Tudor Jones and the number two in the world was Moore Capital. The Premier League also includes Bruce Kovner, D.E. Shaw and Renaissance Technologies. "These firms were the sorts of firms I wanted to create when I was at AHL – a real investment management company. I couldn't persuade my colleagues or the Man management to give me the backing and that dream receded."
With such yearning in his voice it is clear that this particular dream has not died. "I had laid out my plans to management at Man – I began to hire research teams, I was going to have research into equities, set up research laboratories and to acquire the data. But I just couldn't get the backing. When I think of what those companies went on to do …D.E. Shaw, Renaissance look at them now. Then there is Citadel: Ken Griffin is even younger than me! So yes, I do want to create that now."
"Someone I haven't mentioned yet is Brian Draper, who was my guru," says David Harding. "I worked with him over a number of years at AHL and after, and he was my formative influence. He was an older man, and he was a statistician and an actuary. He was much cleverer and much more highly educated than any of A, H or L. He taught me a lot about being in business, about investment management, politics and statistics. He developed a lot of the techniques that lie behind a lot of people's research today. He "wrote the book" for systematic trading."
He continues, "The original AHL system I would describe as "Heath Robinson-like". Yes of course it worked, but it was rather "this bit here does this", and "that bit there does that". There was no overarching principle behind it. Brian was intelligent enough not to dismiss what we were doing as complete nonsense, as the markets weren't efficient and we were able to make money out of them for a reason. He looked at it, and was able to organise it into a proper framework of classical statistics. Sadly Brian died two years ago," baldly states Harding.
Harding continues with analogies of the systems he's been associated with. "What we were using at AHL could have been published, but it would have come out rather like a section of a cookbook, containing Delia Smith's recipe for ginger chicken perhaps. It wouldn't have been a title suitable for a library: for example "Nonparametric Statistical Inference". By contrast, at Winton Capital we not only have the latter title and also next to it: "Understanding robust and exploratory data analysis". We have a body of work here – serious statistical work here." Harding sees it is very significant that he has built on the earlier work to the extent that the foundations are not visible, and the superstructure would be very impressive should visitors be allowed to cast eyes on it.
Harding discloses "the only thing I only ever wanted to be in my time at Man was a quantitative trader, because I think like that. Just like a violinist needs to play a violin, I need to take a quantitative approach to markets. I'm only interested in the maths. What we are doing is finding a way of forecasting the expected returns from futures markets (though it could equally well be CFD's or equities) conditionally based on n explanatory variables, where n could be the ground temperature in Idaho, , the short-term trend, or the rate of growth of the economy. If you had good enough forecasts based on enough explanatory variables, you could theoretically make money every day – bingo, you would become Renaissance (Technologies)!"
Harding in expansive mode can see Winton and ilk as the Everyman of the investment management world. "Calpers or their equivalent don't need to allocate to someone else," he states. "We offer no correlation with other asset classes. Their liabilities are affected by oil prices, interest rates and economic growth and a whole host of other factors in the long term, but we aren't. We are already invested in all the markets around the world so we are a fully diversified investment in our own right. Our risk number is scalable. If Calpers wanted not 80% of the capital in cash (margin to equity ratio of 20%) but 99% of the money in cash then we could do that. We could do overlay programs like Bridgewater, or we could be an overlay on the S&P. We could manage to any risk level that is given to us," suggests Harding. "What I'm saying is that we go to the heart of money management – we are not actually a niche of the money management world."
Harding showed a summary of a piece of research work. The twist on the trading system for the grains sector would have improved the Sharpe ratio on the grains by 20% over the last 3 years and by 15% over the last 10 years. Then comes Harding's problem. It is difficult to get this evaluated. For one thing it is quite valuable, so anyone who sees it could take the idea off to the competition. There could be professional jealousy over the work or attention it receives. There could be a difference of opinion between two very bright people reviewing it, and talking at a level where it can be very hard to mediate. Say then, that a scientific consensus is reached amongst the very narrow group of specialists. The work has to be validated by someone else. "There has to be a peer group review of some rigour," affirms Harding.
Then it has to be built into the whole trading/risk system. Harding looks at the summary sheet and states that the twist delivers only a small improvement. The sub-system for grains system is itself complicated, and the twist may not add enough value to make it worthwhile implementing given the constraints. This trading for grains will affect all the other systems for other markets. There is then an amount of resource that is required to adapt everything else to the new sub-system, so there is a resource allocation decision to be made. Such delicate human issues and finely-balanced research decisions don't get solved by shouting according to Harding – not with teams of extremely bright capable people to deal with. Whilst the research talent is partly motivated by money they are also like to take professional pride in what they do and look for peer recognition. If they don't like the outcomes they can take their talents to a competitor, or set up their own companies. These are the business issues David Harding faces daily.
David Harding and his investors both need patience and faith. Harding says he took on board the necessity of patience from his mentor Brian Draper. He certainly needs it. Firstly he has needed a lot of patience with the generality of potential investors who understand so little of what he does. Then there is the small sub-set of investors who truly could understand what he does at a detailed level. But they will never get the chance, as "looking under the bonnet" at the engine could compromise the edge of advanced statistics and mathematical techniques applied to markets. Usually high-end funds of funds will have an intimate knowledge of how alpha is generated and risk is controlled in a fund where they are invested. But even the most sophisticated hedge fund investors have to act in faith when they invest in Winton funds.
Even having access to positions via a managed account does not disclose either the source of alpha or the forecast correlation matrix that helps control risk, though one could derive margin-to-equity. David Harding has every intention of closing the managed accounts where he can, so even that minor insight will disappear.
What is left is the shared faith and patience. In a drawdown both Winton Capital investment staff and investors have to wait for the systems to generate profitable trades, confident that they always have before, and both believing that the risk controls are appropriate and sufficient. The drawdowns come as night follows day, but so do the positive returns provided the edge is maintained. There you have it: if you believe that the research spend is sufficient in amount and in payoff to maintain the edge, then you should throw your capital in with Winton Capital Management, currently the best performing large-scale CTA in the world. If you do, there is a chance that you are aligning your interests with the first trading operation to join the Premier League in over a decade.
The Sayings of David Harding
"It took the stock market bubble of 1999-2000 to put behavioural finance on the map."
"Our sort of approach to markets is a science. It is an unpublished science, but it is a real one. You could get the thick leather bound volumes of papers on it if there was a willingness to "open the kimono", as the horrible modern expression has it."
"The process of trading our system is like repeatedly drawing different coloured balls from the statisticians apocryphal bag. As we draw out a ball it becomes part of the track record, and we put it back in the bag, but there is no guarantee that the balls will come out in the same order in future."
"We hire PhD's with specific personal qualities because I don't want bright egotistical traders in my organisation: I don't think I can handle them."
"The CTAs are more similar to each other than other hedge fund styles. The processes are similar. Systematic CTAs are bound to have the same trades on, unlike equity long/short which don't have any process at all compared to what we do."
"There are no discrete events in our investment process. We do not put on a trade monitor it and take it off. We stopped doing anything like that 10 years ago."
"If you put in stops and run your profits and trade randomly you make money; and if you put in targets and no stops, and you trade randomly you lose money. So the old saw about cutting losses and running profits has some truth to it."
Three Things You May Not Know about David Harding