Being an investor in an environment dominated by a highly consequential yet uncertain theme is exhausting. Everyone is searching for answers about the impact of AI and acting as if the next lurch higher or lower in the ‘AI trade’ might be the signal they have been waiting for. I’m afraid these market moves are not providing as much information as we might hope.

The pattern currently established in markets, a familiar one whenever a prominent theme emerges, is extreme sensitivity to the next piece of AI-related information. When it arrives, it is either validation of the bull case (“all that CapEx is more than justified!”) or proof that the bubble has finally been pricked. The result is a series of repeated mini-cycles of euphoria and gloom.

There are two things happening here. First, short-term traders are adjusting their positioning based on how they think other investors will react to new information. Whether that information is genuinely meaningful is often irrelevant. Second, fundamentally-oriented investors are struggling to deal with significant uncertainty, overreacting to each fresh piece of news while inevitably feeling the pressure of sharp market moves.

We have an equity market characterised by extreme levels of dispersion, driven by concentrated thematic trading and accompanied by a torrent of commentary eager to explain every move as a dramatic fundamental shift. This creates an incredibly febrile environment. While there is genuine uncertainty about the consequences of AI, it is undoubtedly being amplified by aggressive positioning and exaggerated narratives.

This presents a difficult backdrop for investors, defined by periods of pronounced, concentrated returns and frequent reversals. To have any chance of navigating it successfully, it is essential that we do not treat every mini-cycle of AI pessimism or optimism as the ultimate resolution of the AI story, however comforting that may seem. We are almost certain to experience many more head fakes and false dawns. Each time they occur, the prominent bulls and bears on either side will begin their victory laps once again. Ignore them.

Over time, markets will adapt to reflect the realities of AI’s economic impact. However, that adjustment probably will not happen tomorrow, and we are unlikely to predict either when it arrives or what it will look like.

The good news is that heroic bets on extraordinarily complex technological and economic developments are not a requirement for most investors. Instead, I would focus on three things:

1. Nobody has the answers on AI

In an industry that prizes loud and high-conviction opinions, it can feel as if you either have to listen to aggressive views on AI or hold one of your own. You absolutely do not. Anyone claiming certainty is either wildly overconfident, selling something, or both.

2. Don’t invest as if you have the answer

It is one thing to talk with confidence about how AI will play out; it is another to invest as if you know the outcome. Trying to position your portfolio around a single vision of the future is extraordinarily difficult and entirely unnecessary. The answer to uncertainty is diversification: building resilience to a range of possible outcomes. High-conviction, concentrated investing looks smart. Diversification is smart.

3. The market’s focus should not be your focus

The biggest behavioural risk in periods when markets become enthralled by a single theme is that we start to believe it deserves our attention too. This is a mistake. The market is focused on what is happening right now; our focus should remain on our long-term goals.

–

The impact of AI on markets, society and the economy is highly uncertain, and it is understandable that we look for clues in short-term market movements and become distracted by the hyperbolic narratives that accompany them. However, these bouts of AI hype and doom are likely to be deeply unhelpful. Not only do they contain little useful information, but they also draw our attention towards short-term noise and away from the factors that truly drive long-term investment returns.

==

–

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.

If you ever visit Stockholm, I would recommend a trip to the Vasa Museum, which is home to a grand warship that sank little more than a kilometre into its maiden voyage in August 1628. Aside from marvelling at its staggering level of preservation, the Vasa disaster is a vivid and ever-relevant example of why big decisions can go terribly wrong.

The Vasa was commissioned by King Gustav II Adolf to project military strength and royal grandeur. It was 69 metres long, with three masts and ten sails. It also carried 64 bronze cannons across two gun decks and close to 700 sculptures and ornaments. Looking at its scale and all it carried, it seems a wonder that it ever floated at all.

The design of the ship was the responsibility of Dutch master shipwright Henrik Hybertsson, but he died a year before the ship’s first voyage, which augured poorly for what was to follow.

Before departure, a test took place in which 30 men ran from one side of the deck to the other. The roll caused by this activity was so severe that the test had to be abandoned. Although Vice Admiral Klas Fleming witnessed this, he did not report it to the King, who was out of the country. According to testimony at the subsequent inquiry, Fleming exclaimed:

“If only His Majesty were at home!”

The ship set sail in front of expectant crowds on 10 August 1628. The Vasa almost immediately struggled for stability as winds gusted. Water began to enter through the open lower gunports, and the ship sank. It is estimated that around 30 people died.

There are several technical explanations for the Vasa’s sinking: its tall, heavy upper structure, the relatively small hull beneath the waterline and the open gunports. But the choice to sail despite the warning is also a failure of decision making, and one that seems eerily modern.

How does such a decision get made?

Power and incentives

Most decisions reflect the power and incentives at play. Decisions get made by those with power, based on their goals and desires, and the Vasa was no different. The King was the autocratic decision maker and he wanted a glorious and prestigious ship. Those working for him had limited power and were incentivised to please him, or at least not disappoint him.

To this day, the best way to predict what decisions will get made in almost any situation is to understand who holds the power and what they want.

The consequences of dissent

The prospect of disappointing King Gustav II Adolf seems to have carried considerable weight: even the failed test was not reported to him. We cannot know exactly how those involved weighed the consequences of calling out problems or jeopardising the project, but the personal risks of speaking up may have felt more immediate and specific than the general danger of going ahead.

There is little credit for disasters prevented

One of the inescapable challenges of raising concerns or highlighting risks that could halt a project is that, if you are successful in putting a stop to a bad idea, nobody ever sees the catastrophe your intervention prevented. There is limited credit available for stopping disasters that never transpired because you took action. In many situations there is more personal risk to speaking out than averting dire consequences.

Accountability sinks

No person or group was officially blamed for the Vasa’s sinking. The builders said they had followed the approved design, the designer was dead and blaming the King was politically unthinkable. This illustrates a type of accountability sink (no pun intended), where nobody takes responsibility or acts because nobody is clearly accountable – it is all lost in the system of control. Environments without clear accountability or agency are acutely susceptible to this problem and it remains a familiar feature of modern corporate life.

Expertise is not enough

Sweden employed expert (Dutch) shipbuilders and the level of craftsmanship on the Vasa meant that it withstood 300 years underwater largely intact – quality was not the issue. Rather the problem was one of organisation and decision making. If you place a group of competent people within an ineffective system, then you should expect the system to win.

—-

Walking around the Vasa is an incredible experience, and its conservation is remarkable. It is also remarkable that many of the decision-making problems that led to the great ship’s demise have also been well-preserved for 400 years.

---

---

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.

When people describe the reason for share price moves or shifts in bond yields, it is almost always in fundamental terms. As if every time a new piece of market-relevant information is released, investors are hurriedly updating their valuation models. This gives a pleasing sense of order to the fluctuations of financial markets, but is a poor explanation for what is really happening. Most market moves are about investors guessing what other investors are going to do.

Let’s say there is a weak employment number in the US, and the S&P 500 drops by 1%. Is that a result of investors reassessing the fair value of US equities due to an increasing risk of slower growth? Maybe. But it is more likely to be investors reacting to how they think other investors will respond to this new information.

Markets – particularly over the short term – are much like what was traditionally called a ‘Keynesian Beauty Contest’ or a guessing game where you win not by guessing the right answer, but what you think others will say the right answer is.

Does this behaviour have anything to do with changes in the fundamental worth of an asset? Yes and no. If investors are predicting how other investors will react to new fundamental information, then they indirectly care about fundamentals. However, they don’t have to. If they are just trying to predict other investors’ actions, they care about whatever signals those investors might respond to at that time. These could be relevant to the asset’s long-term value or not at all.

Also, if most people are involved in the same guessing game, then it is not necessary to predict how other investors will react to new information, but what other people in the game expect other people in the game to be doing. (It can get very circular.)

We can easily end up with a market where nobody is paying any attention to fundamentals; everyone just acts as if everyone else is.

While a lot of market activity has little to do with genuine fundamental changes in the value of an asset and even if it does it is indirect. This type of ‘guessing game’ behaviour can, however, impact fundamentals, if we bring in Soros’ concept of reflexivity.

Here significant movements in price (whatever the cause) can affect fundamentals or ‘reality’. Imagine French government bond yields are being pushed higher by investors guessing that other investors will react to a worsening fiscal situation and higher inflation. This investor-led price move can morph into a profound fundamental problem as these higher yields create higher debt service costs and increasing political pressure.

Although they can often feel entirely distinct, asset prices and fundamentals are not entirely independent of each other.

The Fundamentals Will Out

If most fluctuations in markets are not driven by investors changing their views on the fundamental value of an asset, does that mean they don’t matter? I don’t think so, they are important in a variety of different ways and over different horizons:

• To the extent that investors playing the guessing game care about fundamentals (often they don’t) then fundamentals matter even if indirectly.
  
  
• In certain situations reflexivity can take hold which means non-fundamental price moves create fundamental issues.
  
  
• In extreme situations fundamentals tend to rise to prominence  – for example, when a company is in distress its fundamental worth (or lack of worth) will matter a great deal to the price of its debt and equity.
  
  
• Over the long run, fundamentals will likely win out in defining the return of an asset. Although changes in fundamental value may tell us little about an individual short-term price move, fundamentals should be the most important element over long horizons.

When we read fundamental explanations for why markets have behaved in a certain way, it is rarely an accurate reflection of why prices have moved. Over most short horizons it is not fundamental changes in the worth of an asset that matter, but investors guessing how they think other investors will react to new information they currently deem important. At times this might be related to fundamentals, but very often it will not be.

To be successful taking these types of short-term views we need to somehow be able to predict what issues investors are currently focused on, how these issues will play out and how other investors will react to them. This is a tough ask.

For long-term investors, the fact that short-term market moves are largely the result of investors chasing each other around shouldn’t be a major concern. That is not a game we need to play. We just need to avoid being fooled into thinking that every daily or monthly price move reflects a meaningful change in an asset’s fundamental value.

---

---

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.

One of the most common views expressed in financial markets at present seems to be that bonds (particularly the government variety) are no longer an effective diversifier when held alongside equities in a multi-asset portfolio. These arguments are usually accompanied by a chart displaying rising stock and bond correlations, and perhaps some five-year performance numbers showing how poor fixed income returns have been. I think these perspectives get a lot wrong about the purpose of diversification and how bonds are behaving.

Bonds are doing bond things

If I had been told at the start of 2021 that over the coming years there would be several supply-side inflation shocks, substantial fiscal stimulus, inflation consistently above target and a material rise in interest rates – would I have been bearish on bonds? Even as someone who is reticent to make predictions, I am pretty sure I would have been – particularly starting from such unattractive valuations. The performance of bonds – and their changing relationship with equities – has been just what we would have expected given the starting conditions and subsequent environment.

It is odd that an asset class behaving in a way that is entirely consistent with its structural features is a reason to abandon it, or suggest it is not working. Bonds seem to be working just fine. Unfortunately, all asset classes are vulnerable to a particular set of circumstances, and the precise timing of those circumstances is usually well beyond our capabilities. This is why we diversify.

Starting valuations matter

The valuation of bonds matters a huge amount to their expected future returns, and it is also vital to their diversifying properties. Higher starting yields provide strong insulation from losses incurred due to rising yield levels, offer greater scope for yields to fall during downturns and provide better long-term returns in the unknowable future paths that lie ahead of us.

Valuations are not just about headline yields (inflation matters a lot too), but lots of pessimistic takes on bonds seem wilfully oblivious to the fact that higher starting real and nominal yields are a good thing for investors.

Don’t judge an asset class by its (short-term) performance

The powerful narrative around the ineffectiveness of bonds is classic investor behaviour. We take recent returns, build a story to explain them and then extrapolate. Using the same process to describe bonds as broken now would lead us to be incredibly bullish in 2020 – they had performed remarkably well in prior years and exhibited a negative correlation with equities, what was not to like? (Apart from valuations that is, but who cares about those?)

An approach to diversification that relies on the performance profile of an asset class through a particular period or ‘regime’ is inevitably deeply flawed. Diversification is about building a portfolio for a range of future paths that are at best uncertain and at worst entirely unknowable. Making allocation decisions that assume asset classes will replicate their behaviour from one specific (usually recent) instance leads us in the opposite direction of what diversification should be looking to achieve.

Is it just a macro forecast?

A key question about the pessimism surrounding bonds is whether it is driven by a belief that the merits of high quality fixed income as a strong, diversifying asset no longer exist at all, or is it simply a prediction that the future environment will continue to be difficult for bonds? These are easy to confuse but are very different stances.

The first is the ‘bonds are broken’ argument – which contends that bonds will not behave in the way they have historically. This is a bold call and one best tested by asking – how do you think bonds will perform in a growth shock / recession? If the answer is poorly, then someone genuinely believes that bonds have undergone a profound, fundamental change.

The second perspective is just a macro forecast. A belief that bonds will continue to behave like bonds, but that the backdrop will remain challenging – presumably because of growth and inflation dynamics. It is perfectly reasonable to hold this view, but also important to acknowledge how bad we are at making economic forecasts, and size any position accordingly.

—

It is easy to think of diversification as just being about combining assets to hit some risk (usually volatility) target or to smooth the path of returns over time. While these are valid reasons for why we might diversify, at its heart diversification is really about the future being incredibly uncertain. We don’t know how the years ahead will play out and our portfolio should reflect that ignorance. We want to own assets, such as bonds, that we expect to fare well (and not so well) in a range of different environments.

It is possible that the underlying features of bonds have changed significantly and they no longer offer anything valuable to a diversified investor, but it is difficult to find evidence of this. It seems far more likely that overconfident investors are making assumptions about the future of bonds based on recent performance and overly confident economic predictions.

---

---

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.

Imagine if there were some dramatic developments in the production of nuclear power that made it far easier and less costly to produce. It would provide abundant, cheap energy across the world, but with one small drawback. There was a 10% chance that at some point over the next 50 years it would kill all human life.

I think most of us would assess those trade-offs and suggest we hold off.

Yet claims of existential risks arising from progress in artificial intelligence have become part of normal public debate without provoking anything resembling the reaction we might expect. This tells us a lot about how we think about risk.

In AI safety, the risk of catastrophic outcomes has been a consistent topic – often referred to, in a wonderfully anodyne way, as P(doom) – the probability of an AI-caused existential catastrophe. Recently, Evan Hubinger, a safety researcher at Anthropic, said he believed there was a greater than 10% chance of AI killing all humans within the next decade.

Some people working at the frontier of AI believe their field could have the most disastrous consequences imaginable for humanity, yet development continues at extraordinary speed. This seems odd.

I don’t want to make this post about whether these estimates are in any way reasonable, but rather about why we are prone to underplay some risks and wildly overstate others.

From a behavioural perspective, our sense of risk is heavily influenced by how it makes us feel – the emotions it provokes and the scars it leaves. This is sometimes described as the affect heuristic. We respond strongly to risks that are salient, emotive and easy to recall, and can be remarkably complacent about those that aren’t obvious and do not stir us in any way.

While some people are unconcerned about the cataclysmic risks stemming from AI because they believe the claims are preposterous, there is another issue at work: even high probabilities can fail to elicit much of an emotional response when the risk itself feels abstract and remote.

Now think about the earlier example of nuclear power. Some opposition to nuclear power – in a world where it could help reduce climate risk – is likely shaped by the feelings provoked by the concept of nuclear. The very term brings up images of bombs, devastation and reactor meltdowns. The risk seems both high and consequential because those negative associations are so readily available to us.

By contrast, for many of us AI is benign. For the moment, at least, it can seem like little more than something that helps make our everyday lives a little easier. The most affecting negative marker related to AI might be reserved for those who have watched Terminator 2 recently.

What does this mean?

The use of bland probabilities will likely have relatively little impact on how people evaluate the risk from continued AI development. This might only change if there were a significant negative occurrence related to AI – let’s say a prolonged power outage or material data breach – that affected our lives in a meaningful way. Something like that could radically change how the risk is perceived.

How we feel about AI risk is instructive of where investors often go wrong when dealing with risk. We are constantly responding to whatever is vivid and prominent, while easily losing sight of the things that are more abstract or harder to imagine.

We focus on the things that we see, feel and remember – the pain of the last bear market, severe daily price swings, the latest geopolitical flashpoint. These events either leave deep emotional scars that stay with us or make us overreact to risk in the moment. They shape our behaviour.

Not only can this lead us to make poor choices about risks that are nowhere near as consequential as they seem, but it also means that we ignore or forget those that are not obvious.

This is why a difficult quarter in equity markets can feel far more threatening than, for example, the risk that we fail to stay invested over the long term. One of these is a material risk for investors, and the other is probably not; but our tendency is to get it the wrong way round.

We deal with risk in a very human fashion. We learn from emotive experiences and focus on what is happening right in front of our eyes. We judge risks based on how they make us feel.

—

---

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).

Amidst the “panic”, “fear” and “tumult” in government bond markets, it can be difficult to keep a cool head. Short-term trading is easily conflated with longer-term considerations, and the prevailing narratives can quickly become hyperbolic and often apocalyptic. This is deeply unhelpful for most investors – what’s the best way to navigate such situations?

In spells of real or manufactured market stress, it is far too easy to lose sight of our own investment approach and the things that we consider important. Most typically this means long-term, fundamentally-minded investors being dragged into short-term thinking. 

I have found one of the easiest ways to uncover what an investor thinks is actually important (or to get them to focus on it) is to give use the crystal ball test. Ask: if you had to predict the performance of an asset class over a period of time and could see one market or economic variable in the future to assist you, what would it be?

We can use this approach to cut through a lot of the noise engulfing government bonds at the moment by asking something like:

• If you had to predict where 10 year US Treasury yields would be in five years’ time and could use a crystal ball to see one metric in the future, what would it be?

Would it be growth, inflation, the deficit, debt to GDP, the Fed balance sheet – or something else?

This obviously has nothing to do with actually forecasting the future, but is rather a way to focus on the things that should matter to us about the asset class under scrutiny.

It can also prove insightful to invert the question and ask: if you knew X in advance, would it help you predict the performance of the asset class? This is particularly helpful during times of market stress, where investors can veer dramatically from whatever their process should be and instead become wrapped up in whatever the market is obsessing over. (The next Treasury auction is probably not as critical as we think). 

It is amazing how much time market participants spend talking about issues that probably don’t matter to them given their approach and goals, and, even if they did, they couldn’t predict how they would unfold. This problem becomes particularly acute in stressed environments. 

Fundamentally-focused investors who care about quaint things like cash flow and coupons are incredibly vulnerable to getting caught up in the activity of short-term traders, because this is what drives markets in the moment. Finding ways to draw ourselves back to our own process and principles is vital.

The other piece of information that the crystal ball test can elicit is an investor’s real time horizon. If I were asked to predict Treasury yields over one month and over five years, the variables I would want to see in advance would be very different. Factors that might matter over five years, probably won’t tell us much over the next few weeks. (I am not actually sure what I would want to see in my crystal ball if I had to make weekly predictions).

One of the best way to understand an investor’s genuine time horizon is to ask them what they would most like to know in advance, and work back from there. If they want to see the next inflation print, they are probably not a long-term investor.

—-

While we cannot predict the future or forecast asset class behaviour accurately, asking ourselves or someone else what we would do if we could peek into the future can help to get the focus back to what’s important to us and away from lure of market noise.

—

–

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).

At any given point in time, there is a risk in financial markets that is the focus of everyone’s attention. Recently this has been US Treasuries with talk of a ‘fiscal crisis’ and ‘debt sustainability’ increasingly common. This, of course, is not a new issue. In fact, it seems to rise to prominence whenever there is a gap in the schedule of ‘things to worry about’. Rather than go too far into the weeds on this topic, I would rather focus on the dangers of the language that is often used when this subject is discussed, which is often hyperbolic, ambiguous and difficult to verify.

US Treasuries are one of the largest and most important asset classes in the world. Not only are they a critical holding in most portfolios, but they are also fundamental to the functioning of financial markets and the global economy. Their fortunes matter to investors in a profound way; that means that using words like ‘crisis’ or suggesting debt is ‘unsustainable’ has consequences.

While you may truly believe that there is an impending crisis or that US debt is unsustainable, you cannot stop there. The issue is far too significant to make vague, headline-worthy statements with little substance behind them. You need to give more – such as answering these questions:

– What do you mean by a crisis – specifically – what would it look like?

– Crisis suggests something sudden – is it likely to happen soon?

– What would the implications be across financial markets?

– At what point does debt become unsustainable? Is there a threshold?

– The US cannot run out of money to pay its debt, so do you really mean an inflation problem? If so, what are you forecasting will happen to inflation?

– It sounds pretty bad if US debt is unsustainable. What should I do? Not hold bonds? Not hold equities?

– How can we test whether your views are valid?

There are some effective tactics at play when people talk about US Treasuries in a dramatic and imprecise way. Key is exploiting the fact that people’s standard mental model is to think of US government debt in the same way as a household or company – as if at some point they simply won’t have enough money to pay the next coupon.

For the avoidance of doubt, the US cannot run out of money to service its debt like a corporate (because it issues debt in its own currency) and neither will it suffer a mirror of the Eurozone debt crisis (for the same reason).

None of this means that there are not risks here. There clearly are. But if you are going to talk about fiscal sustainability, it is very important to be clear about exactly what you mean. If the issue is that sustained deficits on top of already high levels of debt, and a lack of willingness to address them, will lead to future inflationary or currency problems, then say so and invest accordingly.

In addition to the often incorrect framing of the ‘credit’ risk faced by the US, there is also a remarkable lack of verifiability to the prognostications offered. When someone talks of fiscal sustainability in the US, it is often not clear whether they think that this means that 10-year yields will move 30bps higher, or the entire market will implode. That ambiguity makes accountability difficult.

It is so easy as an investor to talk in bold terms about the fiscal risk in the US. You can take the occasional victory lap when yields rise a little, but then claim the big risk is still to come; it just hasn’t happened yet.

This is incredibly unhelpful for other investors. Not only is it likely to cause a great deal of worry, but there is no way for them to know what to do about it. They are left with the notion that there is a major risk overhanging markets with hugely uncertain impact and timing.

If your view is that deficit and debt levels mean that the term premium should increase because of greater levels of inflation uncertainty, that is a reasonable, implementable and testable investment thesis. It doesn’t need the language of crisis.

If you genuinely believe that US government debt levels are a far more serious problem, then show your working. How does this play out? And what should investors do about it and when?

Espousing views about market risks that are exaggerated, light on detail and difficult to validate is fantastic for attracting attention, but it doesn’t necessarily make for great investment insight. We should think about how we use our language, because it matters for how investors feel and act. 

—

---

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).

When people discuss the benefits of adopting a ‘do nothing’ approach to investing, I am always reminded of the study where – when left alone in a room – a majority of men preferred to electrocute themselves rather than be stuck with only their thoughts for company.

I am a strong believer that most investors would be better off acting rarely, but being human makes this a far harder task than it seems.

If there is evidence that we are prone to do ‘stuff’ even when that ‘stuff’ hurts us, then the chances of us resisting the pull to act in an environment that demands and incentivises investment activity seem remote.

Of course, this is not the case for certain investors who, mercifully, don’t have to pay too much attention to their investments; they can be happily distracted by other things, such as enjoying their life. I often feel that the most successful investors will be those who can make a few sensible decisions at the start and then spend their time on other things.

Either deliberately or inadvertently, this type of investor makes their long-term investment outcomes a consequence of the things they are invested in – primarily the compounded profits of the companies they own – rather than the trading activity they undertake. (One of these approaches tends to work, and one doesn’t.)

The problem is that it is increasingly difficult to invest this way. Not only are we wired to want to act, but technology increasingly demands our attention, and parts of the industry profit from our activity.

For investors, doing more is becoming irresistible.

One of the key lessons from behavioural science is that if you want to encourage a behaviour make it easy, and if you want to discourage it make it hard. Friction matters far more than we think.

It is difficult to think of many ways in which it hasn’t become far easier to make bad investment decisions. We are increasingly encouraged to do things that are either unnecessary, damaging, or dangerous.

How do we encourage investors to do less for longer? 

—

---

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).

There was a recent article in the FT describing how retail investors in Korea had suffered huge losses after a sharp reversal in the performance of AI-favourites such as Samsung and SK Hynix. The severity of the falls led one investor to ask: “When can I get out of this hell?” That such market moves can have an impact of this magnitude is facilitated by innovations such as single stock and leveraged ETFs. This is another example of where innovation can amplify the cost of our behavioural biases.

It is irrefutable that most of us have a range of behavioural traits that are ill-suited to making sensible investment decisions. We want to make money quickly, we naively extrapolate recent trends, we chase performance, we are overconfident in our own abilities, and we ignore risks that are not obvious to us. The critical question is not whether we are susceptible to such destructive tendencies, but rather how easy it is for us to act upon them.

From the perspective of behavioural mistakes, we should care about two aspects: their frequency and their magnitude. The cost of poor investor behaviour is dictated by how often we might make errors and the impact of those errors.

In a world where technology has made trading easy and frictionless, and activity is vigorously encouraged, we should expect the frequency of behavioural mistakes to increase. This is a problem inevitably exacerbated by the sheer amount of financial market noise we are exposed to.

It is not, however, simply the ease with which we can make imprudent choices; it is how consequential they are when they occur.

As well-demonstrated by recent events in Korea, the ability of investors to risk catastrophic losses – through leveraged vehicles, options or even highly concentrated funds – has never been greater.

Innovation has seemingly created an environment where investors will inevitably make more rash decisions and be more vulnerable to disastrous, irrecoverable outcomes from those choices.

It is possible to frame this situation as a case of technological progress in the investment industry inadvertently promoting poor and consequential investor behaviour. This is often not the case. Rather, the purpose of much innovation is explicitly to benefit from clients’ behavioural biases – they are known and exploitable. Typically, retail investors bear the cost, while other parties benefit.

For any type of investment innovation, there are three questions that need to be asked:

1. Why does an investor need it? (There is a big difference here between need and want; just because you can sell it doesn’t mean you should)
   
2. What behaviours does it encourage?
   
3. Who stands to benefit from it?

If the innovation is not encouraging better investor outcomes, net of the behavioural impact, what purpose does it serve?

---

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).

For a multi-asset investor, being underweight equities is a notoriously painful position. They are typically the highest-returning asset available, and reducing your exposure can hurt both in the short and long run. Does this mean you should always avoid doing so?

Not necessarily, but your answer to this question probably says a lot about your investment approach.

Let’s create a simple example to test this. Assume you have a two-asset portfolio with a 20-year time horizon, a growth objective and a balanced risk profile. You can choose between US equities or medium-maturity US inflation-linked bonds, with a starting 60%/40% split.

Is there a level of relative valuation where you would consider moving money from equities and into linkers?

(The focus here is on valuation rather than any macro or tactical reasons for shifting allocations).

I would imagine that a common answer to this question would be “no” – if you have a growth objective then underweighting equities almost never makes sense. But is that really true? What if real yields on the bonds reached 5%, and based on even optimistic assumptions US equities looked priced to deliver 2–3% ahead of inflation?

Constraints on Behaviour

I think the answer to this depends heavily on what type of portfolio you are managing, and what the objectives and constraints are. There are several important considerations:

Do you have a benchmark objective or a return objective?

A traditional benchmark-based objective (let’s say 60% equities and 40% bonds in this instance) makes moving underweight at any time incredibly difficult. Owning less of your highest-returning asset, even if it is expensive, is likely to prove detrimental over the long term. We also know that, over the short run, valuation differentials do not matter much at all, and equities can perform well in any given quarter, irrespective of how they are priced.

Alternatively, if you have a return-focused objective (cash+ or inflation+), the decision-making focus can be more on hitting that target, and the discomfort of holding reduced equity exposure is far less acute. It does not matter so much if equities perform very well and you are underweight, provided you are still achieving your primary goal.

In this way, portfolios with very similar implicit objectives (long-term growth) can behave very differently based on how they are assessed.

Are you focused on delivering growth or meeting liabilities?

If your aim is to deliver long-term growth, then being underweight the asset most likely to provide that may not prove to be a wise move. If, however, you are looking to meet a set of liabilities and can do so with less risk and greater certainty, then moving away from equities when opportunities arise may be prudent.

What are your risk constraints?

One slightly odd constraint might arise if a portfolio is designed to deliver a certain “risk outcome”, usually defined in terms of volatility (a useful but deeply flawed measure). Even if you felt returns from a typically lower-risk asset (bonds) were likely to be higher than equities, it might be difficult to move assets without appearing to take less risk than you are mandated to. (Cue the debate about what risk actually is.)

The point here is not that our constraints entirely determine the approach you might take to allocating away from equities – there are plenty of other factors that should influence your decision – but that they can be influential.

Valuation Mistakes

Away from these constraints there are two other issues – often misunderstood – about equity valuations and asset-allocation decisions that are critical.

Being expensive is not a strong enough reason to underweight equities.

It seems obvious that if equities are expensive and are set to produce lower returns, then being underweight makes sense, but this is not necessarily the case. To justify an underweight on valuation grounds, you need to believe that equities are so expensive that their prospective return is lower than that of competing assets.

Imagine you believe that when equities are fairly valued they are priced to deliver a real return of 5% per annum, but extreme valuations mean that they are now priced at 3% real. Even if you are right about the lower return, you will still underperform if that 3% is higher than the return available from other asset classes.

Where benchmark-relative returns matter, equities have to be priced to deliver returns lower than the replacement asset – not merely be “expensive”- for an underweight based on valuation grounds to work from a performance perspective.

Valuations are largely irrelevant over the short-run, and their impact also wanes over the very long-run.

The general view is that valuations are close to meaningless over the short term and meaningful over the long term. This is only partially true. I would frame it like this:

Short term: Low. Overwhelmed by sentiment and noise.

Medium term (5–15 years): High. Strong influence from starting yield and potential reversion.

Long term (15+ years): Moderate. Positive influence from starting yield, but mean reversion becomes is likely to be overwhelmed by years of compounding earnings growth.

Whereas over the medium term there is the potential twin benefit of valuation mean reversion and an improved starting yield, when the horizon gets very long the impact of valuation mean reversion (if it happens) is likely to become insignificant relative to the contribution from earnings growth.

In fact, if you are a very long-term equity investor and have identified a cheaper, higher-return market, then you do not actually want valuation mean reversion to happen – or at least you want it to happen as late in your holding period as possible.

Why?

Because if you are investing for 20 years and your preferred market has a starting earnings yield of 7% compared with 5% for the alternative, you want to keep reinvesting into that higher earnings yield for many years. If it reverts to 5% immediately, you receive a one-time performance benefit, but that is likely to be marginal compared with the benefit of the higher yield persisting.

(This assumes the 7% is a genuine anomaly rather than fair compensation for lower growth or higher risk. A persistent, uncompensated gap of that kind should be the exception rather than the rule.)

This gets to the heart of the argument as to why, if you have a very long-run horizon, diversified equity exposure will generally be fine even if starting valuations are rich.

Of course, there are exceptions. In the Japanese equity bubble of the 1980s and 1990s, valuations became so extreme and earnings so cyclically elevated, that the earnings growth needed to outrun the correction was far too high. It took decades to right itself. This was, however, an exceptional period.

The sweet spot for a valuation impact is the medium-term, 5–15-year horizon, which should influence any decisions around how valuations might affect your equity allocations, and any decision to move underweight.

–

The question posed in this piece is a complex and difficult one for multi-asset investors, and one inevitably shaped by the experience of a prolonged and pronounced bull market in equities. How you answer it almost certainly says a great deal about your approach to investing and the environment in which you operate.

—

---

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).