Are we normal? For many quarters, I have counseled investors that we are going through extreme market conditions and that patience was the best strategy. As the panic fades in the rear-view mirror and the road ahead looks less bumpy, I stand by the advice. But I don’t need to repeat it.
With U.S. market indices at record or near-record levels and other global markets rallying strongly, we are returning to a period of normalcy. Sure, we will have noisy events, such as Cyprus, Italy or sequestration. Overall, however, markets are calmer and I think better prepared to handle news. The Italian elections, which many saw as a referendum on austerity, rattled markets for a few days. Cyprus, which I do not think is a lead domino for the rest of the EU, too will pass. It will be messy for Cypriot depositors, including more than a few Russian oligarchs, but it will not be a watershed moment for European Monetary Union. Markets will take it in stride.
Yes, then, we are normal – not in probability terms, as we’ll read below, but in terms of market conditions. The markets in the last six months seem able to process news appropriately. I haven’t given up being bullish. I think valuations remain attractive and that the market moves up have more to go. Indeed, this period of calm – with volatility measures down and correlations low – is highly favorable for the bottom-up, fundamental approach we use here. In these environments, we typically get the best from our analysts.
Part of my goal in these letters, in the days before crisis, was to share with you some of the insights coming from our analysts. The co-leaders of our tech team, Brad Slingerlend and Brinton Johns, compiled some intriguing thoughts about portfolio construction in the face of our challenges of successfully predicting the future. I think it is worth sharing and can help you understand your portfolio construction and your investment decisions.
Their paper introduces you to power laws, complex systems and normal distributions and takes us and our fat tails from Lake Wobegon to earthquakes and anthills. Our business is about the future and generating insight, and I have faith that when it comes to equities, we can do that as well as anybody. I can predict with certainty, however, that we will get things wrong too. So our business becomes making the most of being right and minimizing the penalty for being wrong. Brinton and Brad offer a refreshing perspective about both.
Living in Lake Wobegon
When it comes to confidence in our views of the future, most of us live in Garrison Keillor’s fictional Lake Wobegon: We’re all above average. Behavioral scientists, however, love to show us that the facts paint a different picture. Their studies reveal we’re actually quite bad at predicting the future. Perhaps worse, our track records seem to have no bearing on our persistent overconfidence about being right the next time. So while we’re bad at correctly predicting future events, we’re quite good at being overconfident.
To find this in financial markets we have to look no further than track records of economists’ forecasts. When we look at GDP estimates over the past 30 years, we see that the “professionals” are only good at telling us what just happened. They are particularly bad at catching more extreme events.
Bell Curves and Complex Systems
We’re all familiar with the classic bell curve or normal distribution. Bell curves correctly explain many things – especially things that occur under relatively mild conditions, such as the distribution of people’s height, or particles in a room. If we can reasonably estimate the shape of the curve, we can understand the likelihood of a range of outcomes, such as the chance that a person will be between 5'10'' and 6'2'', say. After all, 99% of all phenomena in a normally distributed curve happen within three standard deviations of the mean. Here’s the catch: Normal distribution curves are really bad at explaining phenomena in more extreme conditions, such as those found in complex systems. For example, under a normally distributed curve, the probability of the huge single-day market decline in October 1987, known as Black Monday, is 10-148. That probability is so fantastically small we’d likely have to wait around trillions of years longer than the history of the universe to experience it again. It was an unusual day, but not that rare.
So we have a question to answer: Are financial markets complex systems? By one definition, a complex system is one in which large networks of components with no central control lead to complex behavior, sophisticated information processing and exhibit adaptive learning. Let’s see, sophisticated information processing, adaptive learning, emergent behavior. Sounds like financial markets to us.
Complex systems are incredibly sensitive to relatively small changes and are best explained by power laws. So what’s a power law? Unlike bell curves, power laws tell us that extreme events are massively more likely than we care to predict. In other words, we see numerous data points in the “tails” of the distribution, giving power laws ‘fat tails’ compared with a normal distribution or bell curve. Power laws in nature are quite frequent. Earthquakes, which are the product of complex systems, follow a power law. Given a certain number of earthquakes, we know that most of them will be relatively harmless but more than a few will be catastrophic. We just never know when and where the next one will hit or at what magnitude.
The existence of power laws means that large changes or events are far more likely to occur than what normal distribution curves would imply. In fact the extreme events are not that uncommon. If you wonder why we live through so many three standard deviation events, power laws make intuitive sense; that is, we’ve been trying to fit the wrong ‘normal’ paradigm onto a complex system. The probability of Black Monday under a power law distribution is more like 10-5, or a decent chance we’ll experience that type of event every 100 years or so.
To summarize:
· Humans seem hard-wired to be overconfident.
· Economists often use math based on normal distributions which underestimate the
probability of extreme events.
· Financial systems behave like complex systems and are best explained by power laws;
extreme events are not all that uncommon.
Ants are Lazy?
Most of us would describe ants as industrious and certainly not lazy. Stanford University professor Deborah Gordon offers a different take. She has been studying the same group of ants for the past 30 years and may know more about the behavior of ants than anyone. What she found is surprising: Most of the time about half the colony is just sitting around doing absolutely nothing. Why?
Certainly they could gather much more food if they all pitched in, right? Going back to complex systems in nature we see extreme events happen with some regularity. What if a flash flood wipes out the part of the colony that is out harvesting or destroys the nest? Or inversely, what if someone sets up a picnic nearby. No problem, call out the reserves. Ants, which are a complex system, have adapted to be resilient to extreme events, even though most days it costs them from an optimization perspective. Ants have survived millions of years precisely because they DO NOT optimize around productivity because doing so would have knocked them out long ago. Ants are built for resilience. In public market investing, resilient companies exhibit the ability to adapt and evolve to changing conditions and are able to quickly recover from or capitalize on extreme events.
The Power of Optionality
Power laws are no secret to venture capitalists. They know that the majority of their investments are going to amount to nothing, but they also expect that a few are going to make up for all of their disappointment and then some. The best management, the most sophisticated plans or ample funding do not guarantee success. In fact, often the least plausible startups are the ones that break through. Who knew that Facebook would become such a massive success when MySpace seemed to dominate the market? To hedge against the risk of uncertainty, venture capital firms fund a number of companies. Many times several of these companies might even compete with one another. Furthermore, VC’s are not afraid to pull the plug on a company that isn’t working. Companies that adapt and evolve to deliver a strong value proposition to the user while achieving a platform within their field survive to become larger companies.
Although it may be a bit counterintuitive, it’s because of extreme volatility and unpredictability that it’s possible for a VC to lose more than they win but still make out like a bandit. VC’s have learned that optionality is far more important than their attempts at predicting the future. We believe this logic holds true in public market investing too, but with one essential caveat.
Investors often talk about conviction. Our definition of conviction does not mean correctly predicting the future. It means correctly identifying resilient companies to comprise the body of the portfolio (fewer large position sizes) while opening the portfolio up to as much optionality as possible in the tail (more small position sizes). This allows us to remain indifferent about various paths the future might take while still benefitting from change through optionality.
When we speak of optionality or of option positions, we are not referring to derivatives or even financial instruments. We are talking about the nature of the business. In that sense, option companies are stocks where the risk may be high but the upside to being right can be tremendous. As we discuss below, the power of that upside means the stock does not need to be big in the portfolio to be impactful.
Amazon.com, Ants, Venture Capitalists and the Importance of Avoiding the Middle
Jeff Bezos, CEO of Amazon.com, prides himself on long-term thinking. The team at Amazon regularly thinks about their business 10 years into the future, but they do so in a surprising way. They think about what’s NOT going to change over that time period. For example, while Amazon has no better idea what changes the future may bring over 10 years than anyone else, they can say with some degree of certainty that 10 years from today customers will likely want cheaper products, more selection and faster delivery. This is an important lesson. Resiliency teaches us to plan for the future based on what’s NOT going to change. At the same time, Amazon loves to experiment. Some of these, such as the Kindle or Amazon Web Services, become major successes. Amazon offers a great example of resilience and optionality working together in a complementary manner at the company level.
Shifting the focus back to portfolio construction, there are a few common reasons a portfolio
experiences material harm:
1. Incorrectly identifying resilient companies that turn out to not be resilient at all. We often call these value traps, or perhaps valuation has caused a stable company to become an unstable stock.
2. Placing an option position at the top of the portfolio. An option position that is too big can easily become a fatal mistake. By definition, a portfolio can only hold a few big option positions, so the manager’s tightly held views of the future had better prove correct. However, lots of option positions at the tail of the portfolio do not depend on correct views of the future, nor can any one of those positions kill the performance of the total portfolio if the view turns out dead wrong. It’s the difference between crashing a car into a wall at fifty miles per hour (large option positions going bad) and crashing a car into a wall at one mile per hour fifty times. The former will likely kill us, while the latter is just an annoyance.
As we think about portfolio construction, complex systems teach us to spend less time trying to predict the future, expect extreme events, avoid optimization, focus on resiliency, seek optionality and avoid what’s in between. This leads to a structure that positions the most resilient names as the body of the portfolio with the names offering the most optionality in the tail of the portfolio. Names that offer neither resiliency nor optionality should be cut out. These holdings often comprise much of the middle of the portfolio, where long term underperformance can slip by without the manager ever noticing. Avoiding the middle is perhaps the most difficult lesson to put into practice and it’s where art and science intersect. To take the logic one step further, one could argue that a recipe for sure disaster is to attempt portfolio optimization based on math that does not account for the extreme events we see regularly in complex systems. In fact, that’s not a bad description of modern portfolio theory’s efficient frontier.
Let’s review:
· Resilient companies exhibit the ability to adapt to changing conditions, but do not depend on a vastly different future to succeed.
· Complex systems follow power laws. We should expect extreme events.
· Conviction does not mean correctly predicting the future. It means correctly identifying resilience.
· Optionality in the tail of the portfolio offers potential upside regardless of how the future unfolds.
· Resilience and optionality are good. Optimization is bad. Focus on resiliency in the head of the portfolio, optionality in the tail and cut out the middle.
Suggestions for further reading:
· The Origin of Wealth | Beinhocker
· Complexity: The Emerging Science at the Edge of Order and Chaos | Waldrop
· Antifragile | Taleb
Please consider the charges, risks, expenses and investment objectives carefully before investing. For a prospectus or, if available, a summary prospectus containing this and other information, please call Janus at 877.33JANUS (52687) or download the file from janus.com/info. Read it carefully before you invest or send money.
Investing involves market risk. Investment return and fund share value will fluctuate and it is possible to lose money by investing.
Past performance is no guarantee of future results.
There is no assurance that the investment process will consistently lead to successful investing.
The opinions are those of Jim Goff and Janus research analysts as of April 2013 and are subject to change at any time due to changes in market or economic conditions. The comments should not be construed as a recommendation of individual holdings or market sectors, but as an illustration of broader themes. Past performance is no guarantee of future results.
Statements in this piece that reflect projections or expectations of future financial or economic performance of a mutual fund or strategy and of the markets in general and statements of the Fund’s plans and objectives for future operations are forward-looking statements. Actual results or events may differ materially from those projected, estimated, assumed or anticipated in any such forward-looking statements. Important factors that could result in such differences, in addition to the other factors noted with such forward-looking statements, include general economic conditions such as inflation, recession and interest rates.
Investment products offered are: NOT FDIC-INSURED. MAY LOSE VALUE. NO BANK GUARANTEE.
Janus Distributors LLC (04/13)
FOR MORE INFORMATION CONTACT JANUS
151 Detroit Street, Denver, CO 80206 I 800.668.0434 I www.janus.com
C-0413-37429 07-30-13 188-15-25415 04-13
© Janus Capital Group
