A model agent

Mainstream economics’ failure to predict the 2008 crash undermined the profession’s public credibility. There’s a new game in town that financial risk guru, Rick Bookstaber, has faith in. He shares his belief.

Her Majesty Queen Elizabeth spoke possibly for the entire country when she famously expressed her disbelief in how the economics profession failed to see the banking crash a decade ago. Others did predict it but went unheeded. And while a repeat is oft predicted, there remains no way to see it coming. There may, however be something on the horizon.

Nearly 15 years before the events of 2008, “one of Wall Street’s rocket scientists,” Rick Bookstaber, first encountered an emergent approach to behaviour modelling that had promise. In 1993 Bookstaber – then head of risk management at Morgan Stanley – was introduced to so-called agent-based modelling at a conference in Santa Fe.  He maintains that agent-based modelling provides something closer to the reality in financial markets than anything coming out of existing modelling approaches.

The approach, now commonly applied in traffic management design, resonated with his observations of the car crash that was the 1987 financial crisis: “Once I saw agent based modelling it really struck a chord for me because I had been in the middle of the 1987 crisis. I was doing portfolio insurance and I saw the process and the dynamic as it occurred.

“I realised people don’t optimise. There are different people doing different things, they came into the market with different time frames. It was much more the way a traffic flow turns into a traffic jam.”

He says he “played around building little agent based models, using C programming,” before letting it go. But he continued to be aware that “something was missing with standard methods.” His realisation of the potential power in agent based modelling to reveal the vulnerabilities in a market was at least part of Bookstaber’s rise to guru status in market risk management.

His recent book in which he expounds the potential of agent based modelling in financial markets is titled The end of theory. But its subtitle: financial crisis, the failure of economics, and the sweep of human interaction, arguably conveys more of its intent.

He says he was particularly drawn to the shortfall in conventional modelling when considering the cascade in liquidity he has subsequently highlighted as a characteristic of crisis. That freefall in liquidity, as leveraged players sell in response to a shock, was apparent again, Bookstaber says, during the failure of Long Term Capital Management in 1998 while he was at Salomon. “And there’s the same story; they were highly leveraged and a shock to the market forced them to liquidate. That liquidation led to a drop in liquidity as people were waiting to see what’s going on,” he says.

“The agent based approach and the dynamic that you facilitate with agent based modelling was apparent to me. I more and more realised this was a way to deal with financial crises.”

“He continued to be aware that “something was missing with standard methods.”

Bookstaber has played prominent roles in both the private sector and pubic sector with chief risk officer roles on the buy-side at Moore Capital and Bridgewater, and on the sell-side at Morgan Stanley and Salomon. Over six years to 2015 he was Senior Policy Advisor to the Financial Stability Oversight Council, and then joined the Department of Treasury in the Office of Financial Research, where he developed an agent-based model to assess systemic financial vulnerabilities. He is currently chief risk officer for the University of California’s $100 billion pension and endowment portfolio.

His move from private sector high flyer to regulation and research was, he says, a “sacrifice” but he had profound ambitions: “I really wanted to do it because in I had sort of sketched out how leverage and liquidity could lead to the crisis. Then it occurred.”

And his counsel and books are highly regarded: “I had given testimony five different times about what needed to be done and why we’re having these problems.

“I felt like if I didn’t try the best I could to help reformulate the financial system, then I’d be sitting around for the rest of my life like one of these old codgers saying: Well you know I remember and if they had just done this. So I felt like if I didn’t do it I would just regret not having given my best shot.”

While at the Office of Financial Research, Bookstaber played an instrumental part in drawing up the Volcker Rule – an element of the Frank Dodd Wall Street Reform and Consumer Protection Act which says that bank dealers can’t do proprietary trading when they’re also client facing. It reduces the risk that bank dealers will disrupt the markets through proprietary trading. “There’s an exception for [Volcker]  – the market making exception which I was instrumental in pushing through,” says Bookstaber.

“If they couldn’t do any sort of trading, then they couldn’t make markets – and obviously you need to do that. And the various broker dealers piped in and said you can’t put this in place because it’ll really reduce liquidity in the market. And of course people in Washington were sceptical of that, they said yeah, yeah sure, whatever. But because I’ve been on the other side, I said no you really have to have this in place.

“So we got the market making exception and I proposed various metrics that could make sure, or hope to make sure that the client-facing trading desks were indeed trading to the benefit of clients and weren’t doing proprietary trading on the side.”

It was three years into his time at the Office of Financial Research when he returned his sights on agent based modelling because “we had to have a way of assessing vulnerabilities in the financial system and that was the tool to use.”

“I had given testimony five different times about what needed to be done and why we’re having these problems.”

But the US is still yet to adopt the approach. “I was sort of a lone voice in the wilderness, so I got a lot done. But the dominant player in anything having to do with this type of systemic regulation is the Federal Reserve. Treasury is more like a central switching station for all the different regulatory agencies; they don’t actually do regulation themselves, interestingly enough. And the Fed, has enough PhD economists to staff probably ten universities. They’re very wedded to the standard approaches.

“So the real application of agent based modelling, it really is more occurring in Europe, than in the US.”

Bookstaber says the Bank of England is a leader. He says the OECD too is “interested in these alternative approached to standard economics,” and he recently gave a talk to the organisation in Paris where he introduced the agent based approach.

“If you just look around at the people who are writing in this field, they’re from Italy, from France, Spain, all over Europe. And actually even the Americans – most of them have moved to the Institute for New Economic Thinking.

“I think the reason is that the US is the bastion of neo-classical economics. Everybody has human capital tied into and to suddenly walk away and start agent based modelling is to walk away from your investments in your human capital – all the students that you have who now are writing things to support your theories and so on.

“Whereas in Europe, I don’t feel like there was ever quite the dogmatic acceptance of neo-classical; it was a little more of a free for all and so it’s easier for people to adapt a new approach. In the OECD it seems like they have a little more freedom to try something new because they’re not so wedded to the old approach. Whereas the Federal Reserve is probably just about the last place that can do it in any real way.”

Cultural hurdles have been the focus of much of the criticism of the sector post 2008. While the regulators too have been the target of criticism for weakness in allowing excessive high risk taking that is commonly perceived as being central to the crisis, the culture and values in banking are seen by many as in need of change to avoid a repeat of the crash. Bookstaber is emphatic in his response:

“The notion that you’ll change the culture I think is sort of a fool’s errand, I just can’t see that happening. The people who go into finance are self-selected and they’re not exactly the Mother Theresa’s of the world. It’s not like they’re venal but they’re driven by a profit motive.”

“The notion that you’ll change the culture I think is sort of a fool’s errand.”

Bookstaber, in The end of Theory, posits that complexity is fundamental to the tactics and strategies in markets much as it is in military practice. And increased complexity, he says, takes the financial markets beyond the capacity of currently accepted economics theory to respond.

He says that markets, like the military, stretch complexity beyond even the notion of reflexivity notably highlighted by star investor, George Soros. In a reflexive market, values are based on subjective understanding, not objective reality. Complexity is compounded because subjective interpretations in the market affect actual values which in turn feed further subjectivity. Bookstaber says markets go a stage further. He holds up what he refers to as strategic complexity, “where people deliberately try to create complexity in the system to gain an advantage in that system.

“Regulation’s never going to be perfect and people will always game around it.”

“In the military, that’s the game. And I give examples of that in the book. There’s a famous military strategist, John Boyd in the US, who’s objective was to create an environment that was unstable with the respect to the adversary. That way you could react more quickly than they could because you knew the complexity you were throwing into the system and they didn’t. So it may not be quite as bad as warfare, but there certainly is a component of strategic complexity in the financial system.

“You think of people trying to operate in markets more quickly, you think of the creation of more complex instruments, creating situations where you have differential information because, say, you’re a market maker. All those things help you create this strategic complexity.”

In this environment, Bookstaber says, “there’s going to be people finding ways to alter the environment to their advantage,” and furthermore “the regulators are always gonna be behind the game because it takes so long to discover what’s happened and to put a regulation in place.” Bookstaber adds: “What regulation was necessary or could have helped stem the crisis in ’08,” but he says he knows no alternative to regulation.

“Reducing complexity and reducing leverage, it just wasn’t that hard to do. We just didn’t see it ahead of time and actually we did the opposite. We made the markets freer. What do we need now? Something that helps facilitate liquidity, I think that’s one of the big issues currently.”

Bookstaber does advocate splitting up the banks but says simply making more, smaller entities would be a waste of time. He explains that banks occupy the three chief layers of the finance world: markets, funders and collateral – central counterparties supporting funding. While banks of any size remain in all three layers there will always be a likelihood that a shock to any layer will lead to contagion in others.

Meanwhile the issue remains that regulators are left as effectively moral guardians of a sector where ethical behaviour is rarely seen as part of the job description. Is that the best we can expect? “I think it’s more that regulation’s never going to be perfect and people will always game around it. But regulation has to do the best job it can; it has to be thoughtful.

“It has to try to anticipate what people might be doing but realise that, the best thing a regulator could do is be able to act more quickly and not be so regimented in what they do because the more tightly defined the regulations are and the more slowly they come into place, the easier it is to find ways around it,” says Bookstaber.

“There is a difference in culture from one institution to the other.”

While he sees an alignment in the financial sector with the greater good as desirable, Bookstaber makes the point that the fundamental driver for the sector – to make returns – is not conducive to a focus on the greater good. He says, however, that his sector-spanning career path has revealed differences between different segments of the industry.

“I think it is true, actually that there is a difference in culture from one institution to the other. Some will simply not do something, even if they could get away with it, because it’s just wrong,” he says adding: “I don’t want to name institutions specifically, but I think if you go through my book, it’s pretty apparent that there’s some that simply didn’t have a sense of decency over profit.”

While he names no institutions he is clear about where  the decency is most abundant: “If you go from the banks to the investment management firms to the hedge funds to the sovereign wealth and big pension funds, I think that a focus on the common good tends to exist more in the last two of the group. In fact a lot of them have mission statements that extend beyond generating returns.

Furthermore, he sees a role within a return-making primary motivation, for the pension funds and sovereign wealth funds to act for the common good

Bookstaber says pension funds and sovereign wealth funds are well positioned to provide protection in a time of financial crisis as a buyer of last resort – rather than the government which he sees as likely to act too slowly and always burdened with political considerations. He says those funds, “armed with agent based modelling so they can understand the dynamics that are occurring,” could use their position of holding lots of capital and being unleveraged with long holding periods, to go into markets that are under extreme pressure as liquidity suppliers.

Bookstaber explains: “So if a market is down 30% because people have to sell. And agents are being forced to sell and other agents who normally supply liquidity are not doing it, then I [a pension fund] will take one-tenth or whatever of 1% of my capital and go into that market and buy when it’s 30% down. And because I’m buying when it’s 30% down, I’m supporting the price so it doesn’t go down 40%. And finally the dust settles, I’ve helped dampen the path of the crisis and things go back up to close to normal and I’ve made 30%. So that’s a case where you can have decency and if you had a mission statement that includes market stability, you could be adding value to the market and making returns as you do so.”

“The best thing a regulator could do is be able to act more quickly.”

Were agent based modelling to prove itself to be a model that warrants confidence, has it then the potential to help shield the world from the impact of financial meltdown in ways that economics has failed to do?

“Ultimately, I’m looking at agent based modelling not just as a new paradigm,” says Bookstaber. “And I really feel that it’s not just another model. I feel that it really is a new paradigm of a way to look at economic man, as economic human.”

Leave a Reply

Your email address will not be published. Required fields are marked *