BlueCrest, the $9 billion asset manager, launched a new discretionary equities strategy in July 2013, just after the ‘taper tantrum’. BlueCrest had previously exited the discretionary equities trading space in 2007 but co-founder Michael Platt wanted to revisit the asset class. Platt and BlueCrest management decided to hire their former J.P. Morgan colleague, Christian Dalban, to run the strategy as head of discretionary equities. With BlueCrest providing about $700 million of seed capital for Dalban and his team to manage in the first year, Dalban swiftly started “building the infrastructure, hiring the teams and growing the business.”
Since then the strategy has met its targets: annualizing in low double digits with a mid single-digit volatility, producing a strong Sharpe ratio. Most strikingly, the strategy has already returned approximately 10% net this year, including strong performance in recent months. So far, this has been a challenging year in the discretionary equity space that has seen some of the most venerable and legendary equity hedge fund managers report double-digit losses. Dalban’s strategy, on the other hand, is well ahead of possible benchmarks, which could include multi-strategy or equity market-neutral hedge funds or indices, as well as some equity long-short funds or event-driven funds; though many of these are running a high level of equity beta that, arguably, prevents them from being appropriate comparisons. Dalban does not expect his strategy to be more than 20% net long or net short, on a cash or beta-adjusted basis, and nor does it envisage managers generating more than 20% of returns from factor exposures.
Defining pure alpha
Dalban explains that “most of the strategy’s risk comes from idiosyncratic or stock-specific risk” and this concept goes well beyond the alpha versus beta split. BlueCrest has developed proprietary systems to identify factor exposures, which include country, currency, sector, industry, and styles such as growth, value, momentum, quality and many more. Some of these factors can be viewed and branded as forms of ‘smart beta’ or ‘exotic beta’ but Dalban’s strategy offers neither. Its strict limits on these portfolio biases seem particularly relevant in 2015: at one point in September, some ‘value’ indices (such as Russell 1000 Value) had underperformed some ‘growth’ indices (such as Russell 1000 Growth) by more than 10% for the year to date. Dalban’s strategy had outperformed a naïve value strategy by as much as 20% at that stage!
Whilst some funds are wedded to value or other styles, Dalban aims to avoid any significant factor tilts over time and BlueCrest systems calculate that in excess of 80% of returns are pure alpha.
Beta is intended to be low, but need not be zero. Dalban’s strategy has so far shown a slightly negative correlation to equity indices. Dalban would expect the correlation to be low “as we are trying to capture alpha which should be independent of the market,” he reasons. Dalban also seems open minded about the possibility that some degree of structural bias has caused the negative correlation. “Very strong up days are not good as shorts tend to outperform longs, and in a slightly down market we tend to make money from valuation changes,” he observes. However, Dalban is not certain that a very stressed down market would necessarily be profitable, as his 26 years’ of experience tells him that, “everything could go wrong in that type of market.”
Multi-strategy with specialist managers
Diversification across many dimensions is one bulwark against losses. “We mix many liquid equity market-neutral strategies, with diversity over regions, styles, fundamental and event-driven strategies,” says Dalban. His hope is that some of the strategies will always be performing well, and therefore this multi-manager, multi-strategy approach will be a more consistent alpha engine than a single strategy.
Some multi-strategy funds divide all teams rigidly along geographic or sector lines but Dalban’s teams seem more flexible. The US teams tend to be sector specialists; in Europe there is a mix of sector specialists and generalists, while in Asia, teams tend to be defined by regions such as Greater China, Hong Kong and Japan. Asia is where most of the emerging market exposure lies; Dalban avoids Southeast Asia, Latin America, Africa, Russia and Eastern Europe, because he says “we do not have an edge in those markets sitting in New York and London.” Nearly all of the teams are located in their local region, because Dalban thinks it is “important to meet companies, and analysts on the ground.” Idea generation is expected to be mainly internal, and fundamentally driven, but managers do have access to myriad trade ideas from banks, brokers and specialist salespeople. But these templates are not set in stone. A few of the teams are global, “over time we think Japan could justify sector specialists,” foresees Dalban, and it is also possible that sub-sector specialists might be hired in future. The three main BlueCrest offices for equities are London, New York and Hong Kong; BlueCrest also has offices in Jersey, Geneva, Singapore, Boston, Westport, Toronto, and Sao Paulo.
For the first three quarters of 2015, Europe has made the strongest contribution regionally and this chimes well with Dalban’s experience prior to joining BlueCrest: he launched the Castlegrove European multi-strategy fund, with some ex-J.P. Morgan colleagues, which was subsumed into Israel Englander’s Millennium Partners with Dalban becoming Millenium’s head of Europe. Dalban claims “we have a higher allocation to Europe than our peers, which tend to be US-headquartered and are more US-centric.” Fig.1 shows Europe making up roughly the same amount of VaR as North America for Dalban’s strategy.
Relative-value and event-driven strategies have contributed alongside fundamental long/short. The highest level strategy split is that 70% of capital will typically be in fundamentally driven equity long/short strategies, with 30% in a wide repertoire of event-driven and relative-value strategies including risk arbitrage, spin-offs, rights issues, holding companies, relative-value arbitrage, liquidations, corporate actions, and share-class arbitrage such as ‘A’ versus ‘H’ shares in China. Dalban likes the blend of fundamental and event-driven as he finds good decorrelation between them. Though long/short strategies are much more scalable, Dalban aims to keep the 70/30 split over time as assets grow. A broad-brush distribution of teams is illustrated in Fig.2.
The BlueCrest culture
BlueCrest wants to avoid ‘key-man risk’ and eschews the ‘star manager cult’. The multi-specialist manager model is all part of the BlueCrest culture. BlueCrest has “a very open culture in terms of ease of communication in relation to market colour between credit, rates and equities teams all sitting on the same floor in an open-plan office,” finds Dalban. But at the same time, the other four discretionary strategies traded at BlueCrest all have separate and independent teams, and within Dalban’s own strategy, discretionary equity, there are also more than 30 independent teams. This model follows the BlueCrest model of well incentivized, highly specialized, multi-manager teams given considerable autonomy but subject to tight risk controls. “The strong platform for risk control and performance control benefits both managers and investors, in terms of controlling risk and volatility,” Dalban states. But BlueCrest is not overly bureaucratic. Having once set up his own firm Dalban appreciates that BlueCrest is “much more entrepreneurial” than some of his past employers.
The fee structure on Dalban’s teams appeals to managers with an entrepreneurial mindset. Each team ‘eats what they kill’ and gets a share of their performance, after their share of the infrastructure costs. “We need to offer a competitive deal to attract and retain the best managers,” argues Dalban, though nowadays the managers do not get the entire pay-out at year-end. They must defer part of their bonuses, by investing into vehicles managed by BlueCrest.
Dalban has built a huge resource of nearly 100 people in three regions, focused on approximately 35 independent strategies and he is now limbering up for the long haul. “Once you identify the right people this is a stable business,” he states, adding that “life expectancy of managers can be a lot shorter in directional strategies such as macro, where managers are more likely to hit their stop and a Sharpe ratio of 1 is good.”
The core building block for a sustainable hedge fund business is managers’ risk-adjusted returns. Dalban expects individual manager’s to have strong gross Sharpe ratios above 1.5. He also aims for the equity strategy’s overall net Sharpe ratio to be over 2. If this net Sharpe ratio can be achieved, then the strategy’s volatility target will be consistent with its return target.
Competition for capital
There is competition for capital amongst the approximately 35 teams managed by Dalban, who are all employees of BlueCrest, and work exclusively for BlueCrest. “They are ranked on a variety of performance measures, including raw and risk-adjusted performance. They know where they stand versus their peers, and this helps to keep everyone on their toes,” says Dalban. A manager who is bottom quintile, knows that 80% of his or her fellow managers are outperforming him or her. Perhaps more importantly, the manager also knows why he or she is underperforming because the factor analytics are provided to the managers to review. “We attribute daily performance between factors and alpha and share these reports with managers on a daily basis to give them a clear idea of where their returns are coming from,” Dalban explains, and this is an advantage of BlueCrest’s know-how. He questions whether an individual manager would have the analytical tools to deconstruct returns into component sources.
Capital allocations per manager are typically rebalanced monthly, but this can happen more often. The largest allocations do not simply flow to those who did best last month, it is about quality of the investment process. “Allocations are all about the quality of performance over time, based on our analytical tools and daily interactions with managers, and in this way capital flows to the best investment processes,” Dalban finds. Dalban reckons he needs at least 12 months to assess performance quality and expects turnover of managers to be around 20% over a two-year period. Allocations per manager will naturally vary, but no single manager is expected to receive more than 10% of the total allocated capital.
Balancing manager autonomy with risk controls
BlueCrest may be renowned for tight risk controls but within their limits managers have freedom to think independently, and differently, from one other. Teams can choose to lever their capital budget up to two times – or de-lever it down to zero and sit on cash. This means that even when Dalban’s top-down capital allocations do not change, the strategy’s overall exposure could expand or contract according to managers’ bottom-up confidence. Dalban is comfortable to delegate some degree of discretion over leverage to the managers because “They are living it every day so they are in a better position to predict their own opportunity set, whereas I cannot be a specialist in all sectors or sub-sectors.”
The managers also have some choice over which instruments they use to express their ideas. “A few strategies are volatility and option specific, and those managers who know how to use options can benefit from implementing ideas through derivatives,” Dalban, who was once co-head of global equity derivatives at J.P. Morgan, explains. The use of futures and exchange-traded funds is permitted, but subject to restrictions. “We recognize that managers tend to have more ideas on the long than on the short side, but they still need to find stocks to short,” stresses Dalban.
Single-stock positions are seldom cut at the behest of BlueCrest risk managers. Though this is possible, Dalban feels “we need to maintain a bottom-up process to keep the diversification alive.” Two independent risk managers discuss the limits in a daily dialogue with managers, which tends to emphasize the “shape and size of the portfolio rather than specific stocks.”
It is perhaps because BlueCrest sets clearly defined limits on so many factor exposures, that Dalban feels that a macro overlay, as employed by some multi-strategy funds, could be superfluous. Individual managers, however, may use macro hedges such as index put options, and Dalban likes to see some convexity because he recognizes that even a broadly market-neutral portfolio can be vulnerable to “downside stress from increased correlations or deleveraging by the street.”
Factor risks are allowed in moderation, particularly if they are the consequence of an alpha rationale for trades, and it seems that some judgment is exercised in disentangling factors from investment theses. An example of acceptable factor risk could include second-order effects of a factor, as Dalban explains, “Managers may profit (or suffer a loss) from an oil price crash because their stock picks benefit (or suffer) from a lower oil price. But this would be a second-order effect rather than taking a view on the oil price.”
Still, Dalban reminds us that BlueCrest is constantly monitoring managers to make sure that “stock-specific risk is driving their returns, and that market, sector or style factor risk is typically less than 20%,” he says. Eliminating all factor exposures could end up restricting managers to a pairs trading strategy that is akin to some forms of statistical arbitrage. But Dalban insists BlueCrest does not go this far – “we are very conscious about factor risks but managers are not pairs traders,” he points out.
If some degree of latitude is allowed over factor exposures, each of the teams faces some hard limits. Their largest single name exposure typically works out at 5% of their own gross exposure. Then there are many other constraints, on asset concentration, volatility, and liquidity, with very little small-cap exposure. The equity strategy has marginal currency risk and can allow some, but will neutralize it at a certain level. The way in which the strategy is traded avoids currency exposure in most cases: managers trade in local currency, however any local currency exposures are flattened daily into USD. Factor analysis will pick up second-order exposures through importers and exporters, for instance, though even these indirect currency exposures are small. “The Swiss Franc de-pegging in January was a non-event for the equity strategy,” recalls Dalban.
BlueCrest’s operational, risk and business committees may set other limits and Dalban finds “the scale of the firm is a comfort on operational risk.” As his strategy is entirely invested in listed securities with no hard to price OTC (over the counter) items, the role of the firm’s valuation committee has been limited with respect to the equity strategy.
Risk in August and September 2015
These two months may be a good test of risk management as they were the toughest two months of the past two years for many funds, with some down double digits. But Dalban’s strategy seems to have performed relatively normally. BlueCrest naturally noticedcorrelations increase over August and September 2015, and this had some impact on returns. Dalban’s strategy was up in August, “with alpha quality remaining strong,” Dalban is pleased to say, while September was a month of two halves. While performance was also positive in September, Dalban admits it was a poor month for alpha, as market stresses including volatility in China, and events in Europe, made it more difficult to pick stocks.
Leverage was almost halved over the third quarter of 2015. The fact that this was not a response to rising Value at Risk demonstrates just how different Dalban’s strategy is from many others that run more beta or factor risk. “VaR fell by a similar amount as it comes mostly from idiosyncratic risk,” says Dalban. The fall in leverage was largely down to the managers deciding to shrink their books.
Liquidity and capacity
The equity strategy’s first line of defence is not diversification but liquidity. “Liquidity is more important than any risk factor because a liquid book can always be closed quickly,” says Dalban, who is adamant that he “does not want negative convexity at any price.” Dalban is particularly wary of derivative strategies that can do well for years but involve selling puts, including those on liquidity, explicitly or implicitly – as “this would put the business at risk,” Dalban warns. August 2015 was a poor month for liquidity, but Dalban finds the summer is seasonally poor and does not notice equity market liquidity getting worse. The strategy has very little small-cap exposure, and sets a high bar of $2 billion for defining the boundary between small-caps and mid-caps. Some 80% of portfolio positions are less than one day of average daily volume. In light of the high level of liquidity in the underlying markets Dalban does not believe the strategy is currently capacity constrained.