Monday 11 May 2020

Coronavirus update: From an unknown unknown to a known unknown

On March 17th, I blogged some thoughts on the coronavirus outbreak; its significance; and how I was seeing the outlook. While we are still far from the end of this crisis, at this stage events appear to be playing out largely in line with the analysis and predictions outlined in that article, so I haven't felt much need to pen an update. My views were generally considered wildly optimistic at the time, but recent events suggest it was more a case of investor sentiment being unreasonably bearish than my views being unreasonably optimistic.

I've also been very busy analysing stocks and taking advantage of the market volatility (you don't get given buying opportunities of this caliber very often - somewhere between once a decade and once in a generation - so you absolutely have to make the most of them), so I have been allocating less time to writing/blogging. Given the pervasive impact of covid-19 on economies and markets at the moment, blogging about any other topic has also seemed somewhat superfluous.

Nevertheless, I thought I would add a few thoughts in this post. For me, one of the most predictable and amusing things we have seen since markets bottomed on 23 March has been the widespread dismay, confusion, consternation, and even anger expressed by many market commentators and investors about why markets are rallying in seeming defiance of the fundamentals, and 'decoupling from the economy'. Don't investors understand what's happening out there in the real economy? Haven't they seen the payroll numbers? Don't they know earnings are going to drop a lot? Etc etc.

This reaction demonstrates just how little investors seem to understand about how markets actually work - both on the way down and on the way up. I'm not using hindsight bias here - in mid March I publicly said I thought markets were bottoming on both Twitter and this blog, and on March 25, two days into the recovery, I Tweeted "[t]his is about the time in the recovery where you'll hear a lot of people coming out and saying bear market rallies are common, and we will surely go back down and retest the lows, as effects of shutdown will continue for months etc... No such guarantees I'm afraid. Lows are likely in". I'm aware that calling markets is generally a fool's errand, but I also believe that markets are not equally unpredictable at all times. This was a unique situation, and there was ample evidence sentiment had reached bearish extremes, and that investors had likely significantly overreacted. It was not that hard to predict that markets - for both technical and fundamental reasons - would likely revert to a far more rational assessment of the outlook in due course.

Fully unpacking the various issues at play here is difficult because it requires what I call '3D thinking', which does not lend itself to simple narrative exposition. This is because understanding why A leads to B requires you understand C, but understanding C requires that you already understand how A leads to B, etc. This article was written and re-written, and in the end I gave up and decided to just focus on a few pieces of the puzzle (sorry!).

What I want to focus on in this article is the important and underestimated role 'unknown unknowns' play in market turmoil, and the widely overestimated importance of 'known unknowns'. I also critique the idea markets are 'ignoring the fundamentals' - they are not; they are just paying attention to different fundamentals than many people think they should be, but the same fundamentals the market always pays attention to during recoveries (and correctly so). Finally, I address some of the perils of 'reasoning by analogy' rather than from first principles - something I've blogged about in the past.

Known unknowns vs. unknown unknowns: the real reason markets crash (and don't)

In every major market sell-off in history - and every major buying opportunity - there has always been a pervasive sense that 'life as we know it has changed'. When the downturn is viewed from a comfortable historical distance and with the benefit of hindsight, people invariably look back and say, 'gosh investors were so silly to sell stocks down to those levels; if it was me, I would have been smart and rational enough to know it was a great buying opportunity, and bought when others were selling'.

However, prospectively and in real time, when every new crisis comes along, people invariably react by saying 'yes but this time really is different - we haven't seen anything like this before; this is truly unprecedented'. However, what is overlooked is that every major crisis/sell-off in the past has felt (and been) that way to the investors experiencing them at the time, which is precisely why the sell-offs were so violent in the first place. A sense of disorientation emanating from a lack of historical precedent/easy past analogies to draw on is exactly what creates the pervasive sense of fear necessary to occasion a major crash.

The hindsight observer has the benefit of knowing how things subsequently played out with respect to that specific case/downturn/sell-off, which creates a known precedent which can inform their future judgments, should similar circumstances recur in the future. When they look back on past crises, they therefore have a playbook for how crises of that particular nature work, and are then able to fool themselves into believing they would have been in possession of such a framework at the time, and reacted rationally. However, this perspective overestimates the level of visibility investors had at the time had who did not benefit from such a precedent.

This is the very definition of hindsight bias - overestimating how much was known/knowable at the time in light of new information that subsequently emerges. In the 1970s, an OPEC oil embargo, which tripled oil prices virtually overnight, and lead to double-digit inflation and treasury bond yields, was also unprecedented. The world economy was fundamentally dependent on OPEC oil, and the supply had just been choked off. The world as we knew it had changed. After the dramatic and tragic events of September 11, it was also widely believed that the world had changed forever, and that the old rules may no longer apply. The GFC was similarly 'unprecedented' in many ways. The point is, it is always unprecedented, because it if was not, markets would not go down anywhere near as much. You'd have 10-20% ish corrections - sure - but not the 40%-esque meltdown we saw during 1Q20, which was one of the largest and fastest market crashes in financial history.

The error is in believing that it is known unknowns, including known risks/headwinds to the economy, that drive major market sell offs/panics. It is associated with the well-known bias of 'generals fighting the last war': investors believe the next market downturn will look just like the last or prior downturns, and they believe markets will react in the same way. And knowing how those downturns played out, investors feel confident they will be well positioned to navigate the downturn and profit from other investors being irrationally pessimistic.

But markets don't work like that. The additional comfort the above would-be bargain hunter derives from knowing how similar cycles have played out in the past will also be felt by would-be irrational sellers, who are apt to behave far less irrationally the next time a similar downturn occurs, having already been aware of the risk of its recurrence and prepared accordingly. Cyclicals and financials, for instance, have remained cheap in recent years, despite a long bull market, precisely because investors have feared another recession that looked a lot like past recessions, and thus avoided such names. Known unknowns do not drive major market panics, because the necessary degree of fear, anxiety, confusion, disorientation and outright panic is unlikely to be sufficiently present.

It is instead the emergence of unknown unknowns - new and novel risk factors that investors were not previously aware of and had not factored into their expectations/risk appetites, that drive crashes. A crash occurs because it triggers a rapid and synchronous de-risking of portfolios, as people react to the new, unanticipated risk factor, and reposition their portfolios to reflect the increased degree of uncertainty, including the endogenous uncertainty associated with extreme asset price volatility itself (i.e. regardless of its cause, many investors cannot/will not tolerate extremely high degrees of volatility/price declines, and this creates the feedback loop necessary for a crash to occur - i.e. when selling and falling prices beget yet more selling, as de-risking accelerates).

Investors get scared, and they decide to increase their cash allocations from (say) 5% to 20%, to 'preserve capital' in the face of the newly-uncertain environment, and 'prepare for future opportunities amidst the coming downturn'. They do this because they believe an abundance of caution to now be warranted given what has become an extremely uncertain/risky outlook, and also because they mistakenly believe that because the economic fallout is likely to last quite some time, that markets will also inevitably continue to go down/remain weak for a long time to come as well. Ego, 'future opportunities' will be significant, and cash will allow them to take advantage of them.

The problem with this perspective and behaviour is that it is classic 'first-level thinking', instead of the more desirable second-level thinking necessary in markets (hat tip Howard Marks). What investors fail to understand is that it is precisely the synchronised move to higher cash allocations and a more defensive positioning mirroring their cautious outlook - which they themselves were very much a part of - that caused the market to crash in the first place. If everyone increases cash allocations from 5% to 20% at the same time, markets will crash, regardless of the cause. And that has absolutely been the case, from retail to institutional investors, to insurance companies and other institutions alike (QBE Insurance, for instance, recently came out and said they had "materially de-risked the investment book including exiting all equities, emerging market and high yield debt"; HK Exchanges similarly exited 100% of its equity exposure in March and early April).

The thoughtful market observer would ask the question, so what happens next? Most of these investors don't plan to continue to hold 20% cash indefinitely. They are looking to re-deploy it back into equities at a time they perceive to be more opportune. What they really mean when they say that is 'when the outlook is less uncertain and they feel more comfortable', but what they overlook is that sellers will also feel more comfortable at this point; however, the important practical point is that it means they will be future net buyers of equities. Furthermore, they have already sold as much stock as they want/need to sell in order to feel comfortable with their remaining exposure in the face of what they expect will be considerable economic fallout in the medium term. Given their already very cautious outlook, it is therefore unlikely they will sell a whole lot more in the future.

What has happened at this point is that a previously unknown unknown has now become a known unknown, and consequently is now already factored into investor risk appetite and market positioning, and so it ceases to have much impact on market prices. And this is true regardless of whether the underlying economy is weak or not, because the economy does not drive stocks prices - demand and supply do. At this point, and in contrast to the intuitions most recently-scared investors harbour, a further market crash actually becomes extremely unlikely, and those sitting in cash hoping for more of the same are very likely to have their hopes dashed.

This is why markets almost always bottom well before the real economy, and recover in a manner that confounds most investors. Right when the majority of investors have just finished selling down, raising cash, and positioning themselves cautiously and in preparation for the 'coming downturn' and the 'buying opportunities' sure to emerge therefrom, markets start to rally and the opportunities they had hoped and expected to encounter swiftly disappear. The buying opportunities are not created by the economic downturn per se, but investors preparing for the economic downturn by raising cash.

They are left high and dry holding a bunch of cash. They are confused, and perhaps even angry at the market for behaving so irrationally, and ignoring how bad things are in the real economy. They claim investors are ignoring economic realities. They say the rally must be a dead cat bounce. They say bear market rallies are common and investors are being fooled by it. They say investors are too optimistic on the speed of the recovery. They say it's because investors are overly acclimated to 'buying the dip', etc. They use every excuse they can muster to avoid admitting to themselves the sad reality that they may have sold at the bottom, just like patsies do every bear market. What they are really doing is hoping markets go back down so they have a second chance to buy stocks as cheap as they were recently trading, but with so many cashed-up investors similarly hoping for a further pull back, such an outcome is inherently self-defeating. There is simply too much cash on the sidelines waiting for an opportunity to buy the second dip for markets to go down enough to retest their lows.

There never has been, and never will be, a law of the universe that dictates that stocks will go down just because the economy remains weak, and the actual truth is that markets almost never go down much - if at all - when investors are already cautiously positioned for a known unknown, and are already holding a lot of cash in preparation for a difficult and uncertain future. They go down a lot when people go from being aggressively positioned to defensively positioned en mass, because they were not previously expecting future turmoil/economic stress, and they now are. Once people are defensively positioned, markets are apt to rise because the selling pressure of people moving into a defensive position abates, and there is far more latent buying lying in wait (i.e. all the cash people raised to take advantage of buying opportunities), and too few sellers left.

Every single time there is a recovery from a major economic shock/market sell-off, the same thing happens; people react with the same credulity, and believe the market is ignoring the fundamentals. Throughout the substantial 2009 market recovery from the GFC lows, there was widespread skepticism about the durability of the rally. Didn't people know that the economy is in a mess and it is going to take years to recover from it? Yes, they did, and that was the whole problem. It was no longer an unknown unknown. The economy contracted throughout 2009 and unemployment continued to rise, and yet stocks continued to rally, not only through 2009, but for the next decade.

Is it really accurate to say markets are ignoring the economy?

This brings me to my second critique - the idea that the market is ignoring the 'fundamentals' at the moment. While the cash-is-king bears would love this to be true, the reality is that this is simply not the case. What the market is actually doing is what it always does during market recoveries from crises/crashes - ignoring lagging/co-incident indicators that suggest bad things are happening in the economy that were already widely anticipated to occur during the panic, and instead focusing on the leading indicators which are pointing to an improving medium term trajectory.

Furthermore, even if conditions get every bit as bad as anticipated, but they stabilise at those terrible levels, the degree of uncertainty still declines, as fear of the unknown is replaced with tangible knowledge of how bad things actually will get. If you're convicted of a crime but are awaiting sentencing, while your 'sentence' might be bad, there is also some comfort in knowing exactly what it is, so you can accept and deal with it, and don't have to worry about unlikely worst case outcomes. This is a large cause of the well-known 'sell the rumour, buy the fact' dynamic in markets as well. In markets, people can tolerate things being bad; what they can't tolerate is extreme uncertainty.

Everyone expected the economy to take a massive hit and unemployment to surge on account of the lockdowns in March. Panic about the absolute carnage that would befall global economies on account of widespread government lockdowns is what drove many asset markets down 40% in the space of 4 weeks, and many individual stocks down by as much as 80%. Sell-offs of this magnitude are extremely rare, and do not happen if investors aren't expecting something approaching Armageddon. It is therefore not surprising newsflow confirming that a lot of these things happened in April - notably disastrous US payrolls - has had no additional negative impact on markets.

What has actually been happening over the past six weeks or so of market recovery is that investors have been reacting not to the economic downturn that was already widely expected and priced in, but instead to growing evidence that the duration of the downturn and the breadth of severe impacts may have been meaningfully overestimated. For airlines, the downturn has so far been every bit as bad as feared, but for many other companies, earnings have proven far more resilient than many investors assumed they would. This is because - as I discussed in my original article - in actual fact large parts of the economy have remained relatively unaffected, and we now have unprecedented ability to work and shop online in the modern era. There has also been a 'headline' bias - no one was writing news stories during March about countries and states that are not locking down (except Sweden).

Reporting season has aided in this regard, as a void of corporate silence has been replaced with hard facts about the degree of impact companies are seeing in their business, along with post balance-day trends. Generally speaking, results and recent trends have been less bad than previously feared, and most companies are reporting that the nadir was seen in early April, and a steady recovery has been building thereafter, aided by fiscal support which was significantly underestimated. Commentators not acknowledging this reality have been utterly consumed by confirmation bias.

Furthermore, pulling up a bit, it cannot be ignored that the trend in new infections and deaths has been declining almost everywhere. It is no coincidence that markets bottomed on 23rd March, only two days after new case volume peaked in Italy, after which it has exhibited a very substantially declining trend. This outcome is being mirrored almost everywhere, with varying degrees of lag. Social distancing policies have worked, but it hasn't been just that. Even in nations without hard lockdowns, like Sweden, are seeing similar trends, while many states/countries without large-scale public transportation and/or dense populations have seen case volume and total deaths come in way below the expectations that existed even at the time lockdowns were instituted.

This likely reflects a combination of the overestimation of many mechanisms of community transmission (i.e. countries/states where private commuting occurs instead of the use of mass public transportation seem generally less affected), as well as voluntary behavioural changes and increased sanitation practices in workplaces (face masks, temperature screening, testing, etc), which is something I discussed in my original blog article. Evidence is also emerging that a disproportionate amount of transmission in cases leading to death has been happening within the hospitals themselves (there isn't a lot of social distancing in hospitals), as well as in nursing homes (see John Tepper's excellent article on this - a must read). This is important, because it appears likely that the prevalence of community transmission has been meaningfully overestimated, which will contribute to the spread of covid-19 being managed in less economically destructive ways in the future.

Estimated death rates have also continued to plummet as we have increased testing, due to the highly-predictable 'denominator effects' I discussed in my original article. Initially thought to be 3-5%, death rates have now been revised down to closer to 0.3-0.5%, and even that may prove to be an overestimate. Singapore, where very widespread testing is happening, has seen 18 deaths from 18,000 cases - a death rate of just 0.1%. This has two implications. Firstly, it means the lethality of the virus is an order of magnitude less than previously thought; and secondly, it means the number of people that have already contracted the disease is an order of magnitude more than originally thought as well. That means we are coming off a much higher base of virus penetration, which will make it easier to bring case volume and deaths down, and mean there will be a larger than anticipated degree of herd immunity in place to help moderate the severity of second waves.

To see what people thought a few months ago, one need only look at the hysterical musings of people like John Hempton, who after reading a bunch of academic papers and convincing himself he had a 'substantial edge' over the market in understanding covid-19, was calling for hundreds of thousands of deaths in Australia without radical action, and was publicly calling PM Scott Morrison a mass murderer for not doing enough. Australia's death toll as at 10 May has come in at 97 thus far - three orders of magnitude less than Hempton's expectations. Hempton made money shorting the market in March, but did so because enough people - policymakers and other investors - made the same mistake he did. He was right for the wrong reasons. It was this degree of hysteria that was driving markets in March. People really thought the virus was going to kill 3-5%+ of the people it infected, including policymakers, whose extreme policies largely reflected that error of judgment.

Anyone who has studied statistics 101 should have been alert to the dangers of extrapolating from small, unrepresentative sample sizes - in this case, the death rate emanating from the subset of people sick enough to visit hospitals and get tested, rather than the entire population of people infected with the disease. The latter statistic could only be ascertained by a randomised test of a representative sample of the population, and all such tests to date have revealed much wider prevalence than previously expected (and hence a much lower average severity rate). Some antibody tests suggest as much as 15-20% of the population may have contracted it in parts of Germany and Italy.

However, as is often the case in environments of heightened emotion, people failed to reason rationally, relying too much on anecdote and not enough on hard data/statistics - a common psychological failing of human beings, whose intellect evolved to facilitate complex social interactions, not mathematical/statistical nuance. The world is a big place and the population is genetically diverse. There will always be someone, somewhere, who is young and healthy, and has an atypically severe experience with the disease, whose story can be told via the mass/social media. But we have long known that anecdote is a poor substitute for representative data.

The reality is that as bad as covid-19 is from a public health point of view, the data is not turning out anywhere near as bad as people initially feared, and anyone who does not believe this has important implications for the likely severity and duration of this covid-19 induced economic downturn is kidding themselves. Remember, this downturn has not been caused by covid-19 per se, but our policy reaction to covid-19, which has contained some flawed premises. The crisis is entirely man-made, and can be largely undone just as easily by reversing such policies. And as case numbers and deaths have declined, lockdown policies have already started to be relaxed almost everywhere.

Most companies report that activity levels bottomed in early April and have started to steadily recover since that time, aided also by fiscal stimulus measures starting to be felt. Oil demand has started to recover as commuting demand is slowly returning. Even several consumer financing companies in the US have said that the level of delinquencies has started to decline, after peaking in early April, as consumers begin to receive stimulus checks, and being home-bound, are unable to spend money on travel and entertainment, and so are committing funds to debt reduction instead.

China has seen a veritable V-shaped recovery, with many companies operating in the region saying that their factory production is back to near full capacity, and while end consumer demand has not fully recovered in all places, it is rapidly on the way there, despite a significant demand shock to the country's export sector, as lockdown policies in the rest of the world have come into effect. Consider this slide from Adient's 1Q20 calendar quarter for instance. People that say 'we have no playbook for how the economic recovery will look', but we actually do - just look at China.

During market recoveries from significant sell-offs, markets have never focused on how bad things presently are, but instead trends in the second-derivative and leading indicators, and for good reason: the stabilisation of activity at the bottom, and 'green shoots' of initial recovery, are extremely important because of (1) the way the economy works and the way economic actors behave; and (2) the degree to which it reduces psychological stress/uncertainty.

With respect to #1, if you are a company and you see your revenue falling, you react by cutting opex and capex, and increasingly go into survival mode and seek to preserve cash. If your revenue keeps falling, you will keep cutting opex and capex more, and the problem of course is that your opex and capex is somebody else's revenue. This creates a downward self-sustaining cycle in the broader economy, which if left unchecked can spiral into something resembling the great depression.

This is why signs of stabilisation and initial recovery - even if very modest - are so important. Once corporates see their revenue start to stabilise and tentatively improve, they will stop cutting capex and opex further, and may even very tentatively start to increase it - of the 10 projects they cancelled, maybe they will go ahead with 1 that they were previously very reluctant to cancel. Once the curve turns, the self-sustaining forces on the downside flip into self-sustaining forces driving a recovery. As capex & opex starts to turn very slightly up, corporate revenue starts to improve, and then they start to rehire and restart projects, etc. This is why leading indicators are so important.

The second reason is that the degree of uncertainty for investors radically diminishes. When you enter into a downturn and the economy continues to worsen, investors feel like they are walking into a dark tunnel, and the further they walk in, the darker and scarier it gets. They know, intellectually, that the tunnel has a finite duration, but they don't know how long the tunnel is and how dark it is going to get. So long as it keeps getting darker and darker, anxieties worsen. There is a lack of any sense of hope, and it becomes easier for investors to entertain worst-case scenarios, and start to view them as base case likelihoods. 'What if the tunnel goes on not just for hours, but for days, weeks, months, or even years?!', 'What if the tunnel has no end at all!?'. 'What happens if I'm actually walking around in circles and don't realise it!?" 'What if there are strange creatures or undiscovered species of monsters lurking in this tunnel'!?

However, at some point, a very faint light at the end of the tunnel emerges. At this stage, the psychology changes. Now sure - there is still a hell of a lot of really dark tunnel to walk through, which is a real drag (i.e. the economy is going to remain extremely weak for quite some time yet), but now not only is there a renewed sense of hope, but one can also start to form a more sensible and realistic assessment of how dark it is likely to be and for how long, when before one had no idea. The degree of uncertainty diminishes, which allows investors to more rationally assess and quantify the probable economic damage, and the realistic degree of impact on companies' earnings and balance sheets, and increasingly wild speculation about improbable worst case scenarios also tends to ease up. And owing to the market's propensity to significantly overreact in the short term, in almost all cases, such rational assessments reveal that markets have been massively oversold.

People forget that stocks are long duration assets. If a stock drops 40% from 20x earnings to 12x earnings in a market crash, the market has just subtracted eight years worth of profits from the valuation. If the company comes out and says, yeah earnings were down 50% in 1Q20, and we think we will operate at a small loss in 2Q, before recovering to 75% of normal profitability in 2H20 based on trends we are already seeing in our revenues and costs, with a full recovery likely during 2021-22, it is not hard to see how investors could start to reconsider whether such a dramatic repricing of the stock was truly justified. Commentators that therefore say "but the economy is still so bad"; "but their earnings are going to be way down in 2Q", "but unemployment and the economy are going to remain weak for a while yet", therefore completely miss the point. The types of share price declines previously exhibited required things be much much much worse than that to justify.

Investors will look through short term earnings pressures and price the impact in a rational manner provided some critical threshold of uncertainty is not breached. When uncertainty is high enough, investors panic and sell and sometimes, no price seems too low. However, when uncertainty declines back below this critical threshold, investor behaviour becomes far more rational, and when that happens, markets are apt to sharply rally from deeply oversold levels. That is exactly what we have started to see in recent weeks/months, as the degree of uncertainty has moderated.

The dangers of reasoning by analogy

A final point worth discussing about recent market experience is the danger of reasoning by analogy - something I blogged about last year, and which is a very frequent cause of poor judgment in markets (and in life). One of the reasons so many investors have been paralyzed by fear during this downturn is that they have no way of thinking other than by analogy, and there are no analogies to draw on with respect to how to invest through a pandemic and government lockdown, so they have defaulted to one of two positions - both flawed. Either they say that because this event is completely unprecedented, the amount of uncertainty is so high that no reasonable basis for forming any assessment on the outlook exists; or they default to comparing the downturn to the GFC or even the great depression, because - well - um - stocks and the economy went down a lot then too.

There are multiple problems with this approach. Firstly, an aside on uncertainty: the reality is that the future is always uncertain; it just feels more uncertain to investors at certain times. This is because, as noted, the primary risk in markets is not known unknowns, but unknown unknowns. Six months ago, the outlook felt more certain, but little did we know that covid-19 was about to emerge and wreak havoc with the global economy and financial markets. Risk comes from the fact that unknown unknowns exist in the world, and by definition, such unknowns cannot be foreseen. The outlook therefore might feel more or less uncertain to you, but that is an illusion. And even if the outlook for the economy is in fact objectively more uncertain today than it was six months ago, that does not mean the outlook for markets is more uncertain or riskier than normal, because covid-19 is now a known unknown.

Furthermore, the amount of opportunity that exists in markets is directly proportional to the level of uncertainty or perceived uncertainty that exists, because markets/people hate feeling uncertain. That is because investors/people are not comfortable thinking in terms of probabilities - they like black and white, yes and no answers - something I have also blogged about in the past - and if they can't feel sure they know what is going to happen, they don't feel like they can act. But that is not rational - you can weigh the probabilities, and you must, because there is no other way to operate in markets - at any time. And as a buyer, the more uncertain the outlook the better, because the market systematically underprices assets in an environment of uncertainty. Using a high level of uncertainty as a justification for sitting on one's hands may sound sage and prudent, but it is actually foolish.

But I digress; the more important point I want to make here is that the idea that one cannot form reasonable judgments about the outlook and the probabilities in the absence of analogies or historical precedents is flawed. The truth is, the very best opportunities in markets - whether they be at the market, industry, or individual stock level - always exist in situations where there are novel events or changes happening that render past analogies obsolete. It disorients investors and makes them uncomfortable investing. That is how stocks get very cheap. Great investors are the ones who are original thinkers and are able to reason from first principles without requiring the crutch of analogy. This is something Warren Buffett has obliquely mentioned, when he said he seeks out in successors a very rare trait, which is an ability to think about and anticipate risks that have never happened before. In other words, people that think independently from first principles and don't rely on analogies with the past to form useful judgments.

While I might yet still be proven wrong in my views, my article in March was an example of reasoning from first principles. None of those views were based on analogies with what happened in the past, but instead an analysis of the present situation, and the best evidence that was currently available. While it is possible I simply got lucky with my views, as a general rule I believe it is wrong to believe reasonable judgments cannot be formed in the absence of analogy. Indeed, this is exactly the process I follow with individual stocks. The subject matter is different, but the process the same. Being a good investor is fundamentally about the ability to form reasonable judgments in the face of uncertainty and incomplete evidence. If you can't do that, you will feel compelled to wait until the uncertainty is resolved and all the evidence is in, but by that stage the opportunity will no longer exist. The default to lazy, analogy-laden thinking will not work in novel situations, and it can also lead you astray in other situations where analogies lead to important nuances being missed.

Because no pandemic analogy exists, but people feel they need to have one, a lot of investors have defaulted to analogizing the GFC or even the Great Depression. The extent and duration of stock market decline during these periods has often been used as a guide to what we ought to expect this time. One major brokerage firm published a chart showing how the stock market decline was progressing in time relative to those seen during the Great Depression, with a clear implication that this was in some way useful in assessing the outlook for markets at present. This is completely foolhardy. Anyone can say, for e.g, in the GFC markets went down 50% and it took 18 months to bottom, and so in this downturn markets will go down 50% and take 18 months to bottom. The problem is that the nature of this downturn has almost nothing in common with the GFC at all.

One of the potential consequences of the widespread use of lazy and inappropriate analogies is that people's intuitions on the potential shape and speed of the recovery may be very materially off. I have seen many people argue that oil demand will not recover until 2022-23. I've seen people argue that there will be radical and permanent changes to peoples lifestyles, including people working from home long into the future instead of at the office; and staying home and using online services for everything instead of going out. And yet recently reopened cruise bookings for August has surged 600%. This view reflects a combination of profound myopia and a misapprehension of human nature. The truth is, people are desperate to get out of the damn house, meet with friends and family, resume hobbies, and go on vacations. While hesitations will persist for a while out of concern for health and safety, it is human nature for risk to be normalised over time, and for people to learn to adapt to the presence of risk. Huge numbers of people die every year from dengue fever in Asia. It doesn't keep people paralyzed with fear. They accept the risk and get on with their lives.

In my view, there is no sound basis to automatically assume the recovery will necessarily be as slow and drawn out as the recovery from the GFC. It might be, but it's just as likely to be much faster than people expect, because this is not a financial crisis or even a covid-19 crisis - it's a government policy/lockdown crisis that is man-made, and therefore can just as easily be reversed as it was engineered. This crisis had a simple cause - the government forced people to stay home, where many people couldn't work and couldn't go out and spend money. When they go back to work, it stands to reason that productivity and consumer spending will rapidly revert to near normal levels, as we are already seeing in China.

Indeed, if people insist on using an analogy, the one they should be using is what is happening in the Chinese economy at the moment! That is absolutely the best analogy we have to draw on, because they are about 2 months ahead of the West in terms of the infection curve, the economic lockdown hit, and the ensuing recovery. No doubt people will be inclined to say 'but China is different', and yet they are seeking an analogy, and are using analogies far less instructive in its place.

When people look back at this sell off many years from now - particularly investors that did not live through it - they will say 'this was a very obvious buying opportunity; if it were me, I would have been smart and rational enough to buy, because it was obvious the economic impact was only going to be temporary; next time there is a pandemic that tanks markets 40%, I'm going to buy by the truckload'. The problem is that the next event that causes a 40% market crash will not be a pandemic. The next time there is a pandemic (unless it's materially more virulent and deadly), markets might only fall 10-15% because investors will know the playbook and so it won't be as scary. It will be something else that cause investors to believe 'life as we know it has changed'.