When the Seattle Seahawks won Super Bowl LX, Aden Durde became the first British coach to achieve the feat. As Defensive Coordinator he shared responsibility for marshalling the fearsome defense that was the central pillar of the Seahawks’ accomplishment. Following the game Durde was asked to account for the team’s success and he said: “we keep each other accountable and we have a common goal.” 

While this is a very simple explanation that belies a huge amount of work, it does get to the heart of what a team needs to be successful – and is certainly a necessary but not sufficient condition for a high calibre investment team.

Held to Account

In simple terms we can think of accountability as about an individual being responsible for their actions, but from a team perspective it is more about shared trust.

Think of the NFL Defensive Coordinator, alongside the Head Coach, they define a philosophy, design a scheme and call plays. They expect and trust players to operate within that framework.

Players have to trust both the guidance of their coordinator and coaches, but also their teammates. The success of any individual is heavily reliant on other people doing their job well.

Everyone in the team is accountable for outcomes, but also their individual success is incredibly dependent on everyone else within that team. If you trust the other people in the team then you are happy being accountable; if you don’t then accountability feels like an unfair burden.

Cultures within teams are typically defined by the level of trust. If people don’t trust each other, you can forget having a productive culture.

A Trust Deficit

A recent article written by Adam Butler, which I would recommend reading, argues that trust is an incredibly valuable (and largely invisible) common good that tends to be destroyed by market forces – as it is replaced with systems, controls and metrics, which both serve to erode trust and monetise its absence.

The removal of trust and its replacement with measurement is an undoubted problem. Butler tells the story of a nurse whose care and compassion looks like inefficiency and non-compliance on her hospital’s productivity dashboard.

The inviolable pattern is that information becomes a metric which becomes a control which becomes a behaviour. This always happens. 

We see this dynamic in all domains of life – when a footballer (soccer player) is tracked on the distance they run during a game, they will run about a lot – whether or not it is the right thing to do in the context of winning the game.

Controls also always grow and never contract – the system becomes defined by an absence of trust and builds on itself – each breach needing a new control. This can be maddening but is also extremely rational – trust is ephemeral, and controls and metrics are tangible. If something goes wrong, you need to show where the control is.

Atul Gawande’s compelling ‘Checklist Manifesto’ quickly degrades into reams and reams of checks and constraints – more than defeating his noble initial purpose.

This problem matters a lot for investment teams. Both because trust is integral to a high functioning team, and because financial markets are so incredibly noisy – some metrics and measures are required and useful, but vastly more are wildly ineffective. Not only does that mean that behaviour is driven by flawed measures, but the most valuable aspects of an investment approach are lost because they are unquantifiable in any reasonable way.

If you want to understand how an investment team behaves, you really want to know what it measures, how and why.

Of course, nothing is binary. There is a balance to strike – people can be incredibly vulnerable if they rely entirely on trust without appropriate controls in place, but the extension of that is not removing trust from the equation entirely. Without trust, you don’t have much.

Shared Goals

Alongside trust and accountability, perhaps the defining feature of any effective team is that each member of it is attempting to achieve the same thing. The ideal team structure is one which combines distinctive expertise and skills with a shared goal. Too often teams achieve the first part and fail at the second.

If people in a team do not have the same objectives then it becomes virtually impossible to create an environment of trust and accountability. How can I trust the other individuals in my team, if they are attempting to achieve something different to me? This is not a malign or malevolent mistrust, simply one borne from people following different incentives in an entirely rational fashion.

If you are disagreeing with something I say, is it because of your expertise or your incentives? Without shared goals, there is no real way of telling. (Although it is almost always about incentives.)

Imagine an extreme example where an investment team has portfolio managers incentivised only by outperformance, and a risk team incentivised to limit losses. This would be a crazy arrangement, which entirely ignores the trade-offs involved in any risk/reward decision. Although this might seem extreme, it is probably not far off where many end up.

When we are assessing the strength of an investment team, it is easy to assume that the goals are the same – they all want to generate strong returns, surely? Yet it is so easy, and common, for team members to have skin in different games – someone will live and die by performance, someone who needs to sell, someone who needs to avoid anything blowing up. All rational choices for individuals, but a mess for a coherent team.

One of the key challenges for larger investment teams (and an advantage for boutique firms) is that it is far easier to have goal alignment when there are fewer people involved.

Of course, shared goals are not necessarily a good thing in themselves. They are the best way to encourage aligned and complementary behaviour, but that behaviour can be good or bad depending on what the objective is.

When assessing an investment team it is critical to ask – what are the goals of the team (including anyone who has influence on how it makes decisions) and how are team members aligned with achieving this?

Teams need to pull together and in the right direction.

–

There is a lot that goes into building a defence capable of winning a Super Bowl; Aden Durde’s comment on accountability and common goals speak to the foundation that any effective team needs. We should think about these before anything else. 

—

---

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

All opinions are my own, not that of my employer or anybody else. I am often wrong, and my future self will disagree with my present self at some point. Not investment advice.

Last week saw a heavy and rapid sell-off in a swathe of software-related companies, a dramatic move that was linked to Anthropic’s launch of an AI-driven workplace assistant. These share price falls were characteristic of an environment that has moved from AI optimism to AI uncertainty. This is not simply uncertainty about the efficacy of vast capital expenditure, but the jeopardy surrounding which businesses might suffer material disruption from the next leg of AI development. This doesn’t just mean nervous times for some growth investors, but maybe changes how all investors think about risk.

Ain’t Nothing Like The Real Thing

Although it may sound counter-intuitive, one potential impact of the quick and unpredictable progress of AI is a return to favour of old economy stocks. Much of the past 15 years has been dominated by new economy stocks – best symbolised by US technology companies – at the expense of more traditional industries, such as resources, financials and utilities. The market has favoured companies with high price to book ratios often supported by intangible assets.*

In a world where there is increasing uncertainty over which areas will be most at risk from AI disruption, maybe this phenomenon will change; investors might start to prefer companies where the chance of the business being upended seems significantly lower. It feels easier for AI to do significant damage to the model of a software company than it does to a miner, for example.

AI progression means that certain types of business now hold a wider range of outcomes. Almost certainly some investors will find that the intangible assets of their favoured company were intangible for a reason – they didn’t exist. **

The risk attached to businesses vulnerable to AI with intangible assets and once seemingly impervious moats is likely to be a little higher, as will the return required for holding them. It is tough to be an investor if you are living in fear of the next release from Anthropic or OpenAI.

We may enter an environment where (certainly compared to recent years) investors start to value real things a little more and ephemeral things a little less.

The Best and Worst Time for Active Investors

Every consequential investment theme gets turned into an argument for or against active management – depending on our prior – and AI disruption is no different.

Advocates of active investing will assess the current environment and say something along the lines of:

“AI will create a huge gulf between companies that win and companies that lose. Investors who truly understand businesses and the industries they operate in will thrive.”

Conversely, index fund supporters will say:

“There is no reasonable way of knowing which companies will be the beneficiaries of AI and which will be the victims, this makes concentrated stock picking more dangerous than ever. The cost of being wrong has never been higher. It is far better to own everything, and make sure you hold the winners.”

Who is right? Take your pick. What seems incontrovertible is that certain business models will be negatively impacted – perhaps drastically so – while others will benefit. This will create dispersion and volatility.

—-

The sharp decline in the shares of software companies was not fundamental (yet), it was a case of investors predicting how other investors might react to the latest AI developments. We don’t know how genuine these concerns are – and that is the problem. AI is causing huge uncertainty and it is impossible to know what the consequences will be. There will be continued bouts of volatility and only time will tell what is speculation borne of temporary doubts, and what represents genuine and profound change.

—

* It is certainly fair to say that outside of the US, value as a style has been finding more favour in recent years, but the dominance of the US and the Magnificent 7 sometimes makes this a little hard to notice.

** Over a longer horizon there are clearly broader (and more important) questions about how AI will impact society and economies.

—

---

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

All opinions are my own, not that of my employer or anybody else. I am often wrong, and my future self will disagree with my present self at some point. Not investment advice.

There have been some quite significant moves in asset prices in recent times. The dramatic increase in the price of gold over the past year has probably not escaped your attention, but it is not the rise in the yellow metal that is most interesting to me, rather the behavioural implications. How does extreme positive performance from any asset class or fund make us think and act, and why?

The behavioural drivers

Availability: The availability heuristic exerts a huge influence. This is the mental shortcut where we judge something not based on the weight of evidence, but by how prominent and frequent it is. When an asset class keeps hitting new highs, we conflate the volume of ‘evidence’ with its strength. There might be lots of articles, but if they are all saying the same thing, they are far from independent reasons for us to believe the performance will continue. In these situations, availability also leads us to understate risks – because we are not seeing the downside risks, we are likely to severely neglect them.

Stories: Humans understand the world through stories, and when an asset class is on an unprecedented run of good performance, plenty of stories will be told. We will treat these stories not just as explanations as to why something has occurred, but as a prediction of its persistence. We are inherently uncomfortable with things happening without a clear explanation, so to escape this dissonance we look for convincing explanations so that the world makes sense.

Extrapolation: The combination of high-profile performance and persuasive stories inevitably leads to extrapolation – the trenchant belief that what has been happening will persist into the future. The stronger and more sustained the high returns of an asset class, the more we struggle to see anything to stop it.

Fear of Missing Out: We are relative creatures. Our feelings about most things are framed by how it compares to something else (often other people). Two things happen in this regard when an asset is producing strong, high-profile performance. Those of us that don’t have exposure to the asset class will envy those that do, and those of us that do will wish they held more.

Feedback loops

This combination of behavioural factors creates powerful self-reinforcing feedback loops that can serve to further boost an asset’s performance. It works something like this:

Stage One: High returns from an asset class receive significant attention.

Stage Two: Stories are formed to explain the performance; these stories are prominent and convincing because performance is strong.

Stage Three: The combination of high returns and compelling justifications increase the belief that the trends will persist.

Stage Four: More investors are drawn into the asset class.

Stage Five: Returns are boosted by increasing investor appetite.

And so it goes on, until it doesn’t.

These feedback loops are driven by changes in investor behaviour and sentiment, and impact all asset classes over the short-run. For assets that generate cash flows, over time the phenomenon will be outweighed by the gravitational pull of valuations and fundamentals (at some point someone might say “maybe 100x earnings is a little too rich”). This doesn’t occur for ‘belief assets‘ that are not tethered to fundamentals (such as gold or crypto), which makes the range of potential outcomes incredibly wide.

Thresholds

Another fascinating aspect of what happens when the price of an asset rises substantially is at what point do different types of investors become involved (and when might they withdraw). This is clearly a complex topic, but the best model probably comes from Sociology and the work of Mark Granovetter. He explored the subject of crowd psychology and, in particular, how people in a group will have different thresholds for engaging in collective action.

For example, in a group of protestors, there might be some individuals who have an incredibly low threshold for engaging in disorder, whereas others have an incredibly high threshold. Those in the latter group won’t start throwing bricks until almost everybody else has. Granovetter’s argument is that the composition of a group and how behaviour cascades through it will be critical to how it acts.

In a financial market context, we have a huge group of potential investors all with different thresholds for participating in the ascent of an asset class. Day traders, momentum traders, macro hedge funds, all the way to long-term valuation-driven investors. The first group have a very low threshold for engagement, whereas the latter group might only invest when the pressure not to becomes too great (perhaps their job depends on it). The longer strong performance persists, the greater the chance that those high-threshold investors get drawn in.

This threshold model also matters for the reversal of extremely strong performance trends – the key question becoming the reverse: what is the threshold for investors to exit an asset class if returns start to deteriorate? This is what is typically meant when people talk of the dangers of speculative or ‘tourist’ money. Money with a low threshold for exit can create sharp and severe downside risks.

Although the overarching behavioural patterns of investors are similar, their motivations and strategies will be different, and that matters.

–

Extremely positive performance from an asset class can be caused by and encourage extreme behaviour. The more prominent the unusually high returns are, the more important it is to reflect on what might be driving our own decision making.

—–

---

My first book has been published. The Intelligent Fund Investor explores the beliefs and behaviours that lead investors astray, and shows how we can make better decisions. You can get a copy here (UK) or here (US).

All opinions are my own, not that of my employer or anybody else. I am often wrong, and my future self will disagree with my present self at some point. Not investment advice.