Beyoncé, Ego (2008)

The recent publication of the report of the Chilcot Inquiry into the invasion of Iraq in 2003 has been the subject of intense media attention in the UK, at least until it was shunted down the news headlines by continuing fall-out from the outcome of the Brexit referendum. Among other things, there has been a focus on the shortcomings of the workings of government revealed by the report, in particular the handling of intelligence about Saddam Hussein's weapons of mass destruction, and the process of decision-making by the Prime Minister Tony Blair and his Cabinet of government ministers. The report demonstrates the importance of organisational structure for processing and acting on information. It shows how the dysfunctional collective responsibility of the Cabinet system can lead to everyone and no one being responsible for decisions, and how the hierarchy of Cabinet committees results in a lack of accountability. It also shows how assessments of whether Iraq continued to possess nuclear and biological weapons were flawed, suffering from the presumption that Iraq was trying to conceal its weapons programmes, with the result that innocent actions were interpreted as indicative of deceit and failures to unearth evidence were taken as proof of concealment. The result of these failures was an unnecessary war and a failure to plan for the longer-term aftermath, the consequences of which continue to be felt.

In certain respects, the Iraq affair can be seen as a variation of the Abilene Paradox, which describes how a family of four in Coleman, Texas decide one hot, dusty afternoon to go on a four-hour round trip for dinner in Abilene 53 miles to the north. Although each member of the family individually agreed to the decision to go on the journey, it turns out afterwards that none of them had wanted to go. It is an illustration of the failure of decision-making in poorly functioning groups, in this case due to the pressures of social conformity. It is a form of groupthink, other examples of which are the 1961 Bay of Pigs Invasion and the pre-launch decision-making which led to the explosion of the space shuttle Challenger in 1986.

The Iraq affair echoes an earlier chapter in the history of British Intelligence, the failure to predict the downfall of the Shah of Iran in 1979, despite Britain having a significant diplomatic and commercial presence in the country which was of major importance to her interests. A report by the Foreign Office in the immediate aftermath of the Iranian Revolution showed that the Tehran Embassy's research efforts had been directed by London at the wrong area, with a focus on commercial information rather than political developments, and suffered from a failure to get out and tour the country to gather different perspectives. The remote decision-making about the allocation of resources within the Embassy, and the flawed processes for the exchange of views between the Foreign Office in London and the Embassy in Tehran were identified among the ultimate causes of the failure.

All of these examples illustrate the importance of organisational structure to support individuals to gather and process information and then make decisions which generate the best outcome. This applies equally in the investment world as in the political and intelligence spheres. Research has shown that there is a strong relationship between organisational structure and investment performance, as it helps match behaviour to process and creates an environment in which practice reflects policy: a study of 2,000 equity mutual funds concluded that the organisation explains about 70 per cent of the difference between fund results and the manager was only 30 per cent.

At Hosking Partners we took the opportunity presented by a new beginning to design the structure of our firm in order to serve our investment approach and to encourage behaviour which will give us an edge in delivering outperformance. We hope too that our structure should help us not only to avoid falling for psychological traps, but also to exploit occasions when we see others fall into them. Stripped to its essentials, Hosking Partners is built on a multi-counsellor system of global generalists, in a flat structure which has limited central direction and no analyst layer, (1) working within a set of broad investment restrictions, with each portfolio manager implementing a diversified portfolio, each of which are aggregated in a single global equities product. Our business is focused on generating performance fees measured over a long time period, and our ownership takes the form of a partnership which is hopefully forever.

The multi-counsellor structure means that each portfolio manager at Hosking Partners has complete autonomy to manage his own portfolio, each of which are then aggregated into the overall portfolio. It addresses the challenge of assembling a team in a way which gets the benefit of collective wisdom while ensuring knowledge belonging to individuals is not lost, harnessing the diversity of the team members in the most efficient manner. Research has shown that team-managed funds have higher risk-adjusted returns than their single-managed peers, with teams adding up to 30-40 bps per year to gross fund performance.

Key to the success of a team, and as important as individual ability, is cognitive diversity. As Michael Mauboussin has said, being different is as important as being good; alpha lies where the set of good ideas intersects the set of different ideas. Cognitive diversity takes the form of the heterogeneous perspectives, rules of thumb and interpretations which each multi-counsellor brings to the firm. At Hosking Partners, it is taken a step further by each portfolio manager also being a global generalist, meaning that we are each a jack of all trades, applying the general to the specific. And in a final twist, there is no bench of analysts to support the portfolio managers, so each of us has to filter the investable universe for ourselves.

The absence of specialisation is because the information advantage produces diminishing returns: more information is bad, not good, as it encourages the tendency to focus where more information is available. It also leads to confirmation bias, the tendency to search for and interpret information as confirmation of one's existing beliefs: Tony Blair and his hunt for the weapons of mass destruction comes to mind. There is a further advantage to having no analysts: if there is no information advantage, then the advantage is likely to be either analytical or behavioural, involving knowledge which is soft rather than hard. Soft knowledge is harder to transmit than hard knowledge, as it requires persuasion of one by another, so a hierarchical structure involving analysts reduces the incentive to gather it. A multi-counsellor working alone, on the other hand, finds it easier to translate research ideas into investment decisions.

To take advantage of the absence of analysts, Hosking Partners' multi-counsellors have a broadly unconstrained set of investment restrictions, allowing us to exploit the widest opportunity set, with as few limits as reasonably possible on how we invest. This means we can go where the fish are, rather than fishing with the fishermen. It supports a contrarian approach, taking advantage of others' crowding, which is simply a case of diversity breaking down. Often the most attractive opportunities are to be found in the gaps between silos of information: as the market crowds inside easily defined categories, we can exploit the value to be found in the interstices, whether in terms of sector, geography, size, style, liquidity, time horizon or even asset class. It should also mean that we have a relatively high active share, which measures how much the holdings in our portfolio diverge from the index. Of course, we can only outperform against our benchmark to the extent we are different.

Our unconstrained approach is only made possible by having a diversified portfolio with a large number of stocks. Hosking Partners' multi-counsellors each have between 100 and 150 companies in their individual portfolios, which aggregate to over 400 at the overall level. If we are to avoid specialisation, we need to compensate through diversity: in the absence of an information advantage, we would rather be broadly right about many things than hope to be absolutely correct about a smaller number. Conversely, the diversified portfolio allows us to accommodate riskier ideas which would otherwise find it difficult to find their way into portfolios with fewer holdings. It also allows much greater scale to be achieved than with a smaller number of holdings, deferring for longer the trade-off between alpha and assets.

The diversified portfolio also removes the need for management control of the multi-counsellors, whether to control risk or to aggregate decision-making. Reflecting the absence of an analyst layer below, there is also no hierarchy above. Research has shown that hierarchy is bad for diversity as it reduces incentives to take risk and increases the tendency to herd more. Studies have even shown that for US mutual funds an additional layer in the hierarchical structure reduces the average performance by 24 basis points per month! Instead, a flat structure promotes ad hoc sharing of information via shared company meetings, research notes, as well as informal interaction in our open plan office. There is no formal sharing of information beyond our whiteboard on which each multi-counsellor notifies his trading decisions. The purpose behind this organisational approach is to avoid the distortions caused by our natural propensity to imitate each other: this is similar to preference falsification, the phenomenon where social pressure induces people to make public statements that are contrary to their private beliefs.

The automatic aggregation of each multi-counsellor's portfolio into one works because Hosking Partners has just a single product strategy, and all our clients therefore have exposure to the same underlying portfolio containing our best ideas. Capital allocation and stock selection are combined at the same level in a dynamic process which does not require an external decision-maker. Form follows function in a way which is only possible because our strategy is investment-driven rather than marketing-led. The efficiency thus created allows portfolio managers to spend more time thinking about investment in order to generate performance, while the scalability of the structure allows the benefits of growth in assets to be shared with clients in the form of a lower management fee. This last point is of particular importance in a market environment of low nominal returns.

Supporting all this is the partnership structure of our firm. In simple terms, this means that Hosking Partners' organisation is designed and protected by those of us who have to make it work (le patron mange ici), and because the profits of the firm go to those who work in the firm rather than to outsiders, we are also the ones to benefit or suffer directly according to its success. This should contribute towards the achievement of that elusive goal of aligning our interests with those of our clients. The same applies at the level of incentives, where a balance has been sought between equity participation and discretionary bonus awards on the basis of longer-term contribution, reflecting the rolling five-year nature of the performance fee mechanism. This should help address the Ringelmann effect, the tendency observed by French agricultural engineer Maximilien Ringelmann for individual members of a group to become increasingly less productive as the size of their group increases, illustrated by team members pulling on a rope with less effort than when asked to do so as individuals. Finally, we have thought hard about partners' exit, drawing on lessons from other firms with the purpose of avoiding short-term pressures that are at odds with the longer-term investment horizon and the perpetual timeframe of the partnership.

Nothing we have done has eliminated the likelihood of individual failures of intelligence by Hosking Partners' multi-counsellors. Hopefully, however, we have designed a structure which is fit for purpose and even “special”, if not the special purpose Beyoncé has in mind, and which will be able to withstand the shocks experienced in the markets today and in the years ahead while delivering scaleable alpha.