When I sat down to write the four previous Annual Reports for the years 2004–2007, I always had to think hard about what would be the main theme of each report. Things were going very well, and not much was really happening in the financial world. I did not encounter that problem when I started writing the 2008 Annual Report.
The year 2008 was full of unprecedented events and wild market movements. Many books will surely be written about the 2008 market crash, and I will not try to analyse it here too much. Instead, I will focus on the two most important practical questions that shareholders ask:
1. How is it possible that our fund, which for almost four years had a performance that put it among the top 10% of the global equity funds, could have had such a bad performance in the second half of 2008? 2. What can and will be done so that it does not happen again? Let us go straight to the first question. The Fund’s performance can be split into three parts: market performance, fund leverage, and stock selection.
We invest in equities in various parts of the world and, although we hold just a few stocks (30 at the end of 2008) that represent only a tiny proportion of the world equity markets, their price movements, and especially in the short run, are to a large degree influenced by the movement of the world markets. The year 2008 was one of the worst years ever for stock markets. The global market as measured by the MSCI World Equity Index declined by about 40%. Some parts of the world performed even much worse. Europe, for example, was down by 44% and the Czech market by 52%. The second half of the year brought most of the decline and culminated in a brutal crash across all markets in October.
World market performance was a very strong headwind for all equity investors. It is clear that in this environment it is very difficult to make money in equities. This statement does not need much elaboration. Instead, we should ask ourselves whether it was possible to anticipate the crash and avoid the consequences. Many investors claim today that they did foresee the crash coming, but in reality very few anticipated it. Indeed, there are literally thousands of analysts, market strategists and investors making public predictions all the time and it is thus logical that some of them will be right regardless of what happens. No one knows up front, however, who will be right. Since market crashes are actually quite rare and happen maybe only once or twice in an analyst’s professional career, almost no one is in fact able during his or her lifetime to establish a track record as a successful predictor of market crashes.
Let us define a market crash as a sudden (in less than 1 month) broad market decline of at least 30%. How likely is it for something like that to happen? Standard financial theory tries to estimate the probability of similar events using statistical models based on historical market data. The main approach is to look at past market volatility and then estimate the likelihood for market movements of different sizes. For example, a 10% one-day drop in the US market happened on average once in every 4 years, or once in 1,000 trading days. If we consider the 30% crash to be a series of three daily drops of 10% each, then the probability of this happening is once in 4 million years. Could we predict systematically and successfully something that was statistically not supposed to happen in several million years? Based on this theoretical approach, certainly not.
Now you may suggest that we should have considered even the remote possibility of something like this happening and have acted preventively. Well, we had considered that. In fact, we think about possible events and outcomes all the time. Every day, and in 2008 also through many sleepless nights. Acting preventively is not without costs, however. If you expect a market crash every year and you are wrong nine times and correct once, is it better or worse to act all the time as if the crash is coming?
In spring of 2008, we actually assigned a significant probability of a 20% market decline before the end of 2008. We nevertheless decided that we did not need to take any major action since we believed that our portfolio was of good quality yet fundamentally cheap and that it would perform significantly better than the market. What surprised us most was not the market crash but the relative underperformance of our portfolio. More about this later.
Let us go back again for a while to the question of market crashes. We said earlier that it was not realistically possible to predict the October market crash. Nonetheless, we are not happy with this conclusion and have done a lot of work on this since last October.
First of all, we have looked again at the history. History tells us that there were two crashes in the last 21 years that were statistically not supposed to happen in 4 million years. The one in October 2008 and the one in October 1987. In the second instance, the US market fell by 30.7% in 4 days. Then there were 10 market corrections of 12–23% in 1929–1933 lasting 1–8 days. These too, we can say, were extremely unlikely to happen. Thus, we found that there were several market events during the last 100 years that were statistically not supposed to happen in millions or in some cases hundreds of millions years. Was there something wrong with the theory?
In the world of Modern Portfolio Theory and the Efficient Market Hypothesis, market returns are distributed according to the Gaussian distribution.
The curve has a famous bell shape, and extreme events are defined by multiples of standard deviations from the mean. Two-thirds of all outcomes fall within one standard deviation from the mean and the farther the outcome is from the mean the less likely it is that it would happen. For example, the drop by 22.6% in the Dow Jones index on 19 October 1987 was more than 6 standard deviations from the mean and should happen only once in more than 500 million years.
October 2008 showed clearly what is wrong with the theory. The whole theory is based on the assumption that individual events are independent of one another. For example, if you start tossing a coin, it never matters what the previous outcomes were, as the probability of the next outcome is always 50:50. It is independent from the previous outcomes. But this ironclad rule of probability does not at all apply to the stock market. In the stock market, the outcome of the next event (tomorrow) is not independent of the previous outcome (today). During one week in October 2008, world markets fell by 25–30%. It was a sequence of 5 consecutive daily drops of several percent each. When all the markets dropped on Monday, investors did not start Tuesday with new brains and no memory. They were full of panic and fear, and many investors were forced to sell because of margin calls. It was very clear that the outcome of Tuesday was not at all independent of what had happened on Monday. This continued for several more days until for some reason the negative feedback was broken.
We have always criticised the Efficient Market Theory, and it was again proven wrong. Although this gives us some satisfaction, it will not help us to predict the next market crash. What it might help us to do, however, is to buy protection against the next market crash. In the future, we will consider to purchase (when appropriate) insurance against a large market decline. Most likely, that would be in the form of put options that are far out of the money. Because most of the financial world is still using the Efficient Market Theory, such options used to be quite cheap at the times when they give the most protection.
Our investment strategy can be summarised as follows. Vltava Fund is a global long/short equity fund. “Global” means that we do not have a regional focus or restrictions. We are searching around the globe for the best investment opportunities. “Long/short” means that we use both long and short positions. On the long side, we have a selection of individual companies. On the short side, we have a broad market index (or indices). This allows us to be less dependent on market movements and more dependent on our own stock selection. Our long positions are usually larger than our short positions, meaning we are in a net long position. If we are able to find enough good and low-priced investment opportunities, we are fully invested and our long positions are approximately twice as large as our short positions. If we do not for some time see good investment opportunities, we are not ashamed to hold cash and our long position is somewhere between being equal to our short position (so-called “market neutral”) and being twice as big as our short position.
The sum of long and short positions divided by the capital gives the gross leverage, and the difference between the long and short positions divided by the capital gives the net leverage. When we are fully invested, our gross leverage is about 3 and our net leverage is about 1.
The success of our investment strategy depends on the following two conditions:
1. That we are able to identify stocks that in the long run perform on average better than the market.
2. The ability to hold our investments for as long as we like.
The first condition is self-explanatory and does not require further comment. If we do not have the ability to find stocks that over time perform better than the market, our strategy would not make sense and would have to be changed. We believe we have that ability to pick long-term outperformers.
The second condition however is more problematic. When we buy a stock, we do so because we believe that it is significantly undervalued and has potential to go up a lot, often by more than 100% over the next 3–5 years. The price trajectory from here to there can be very unpredictable. It can go up next week or go down a lot first or do nothing for a relatively long time, but, so long as the fundamental value of the company continues to rise, these gyrations never bother us. We are patient long-term investors and can hold the stock for a long time and wait for its price to move up closer to its fundamental value. Provided that we were able to hold our stocks as long as we liked, we were happy.
Then came October 2008, and markets started falling very rapidly. Many stocks were punished by forced selling from scores of investors who had been investing borrowed money, and the stocks literally collapsed. For a while, this did not matter to us because it was clear to us that panic and fear were taking over the market while the fundamental values of most of our companies were influenced much less than the stock prices were suggesting. Since we never borrowed cash, we thought we would just sit it out and wait for the storm to pass.
But as the market decline continued, our gross leverage kept rising. By definition, when a fund’s NAV falls, its gross leverage rises. It eventually went up so much that we were no longer ready to let it grow further, because with it the risk to the whole portfolio was increasing. We decided to bring it under control in order to keep the risk in check. In practice, this meant that we sold part of the long side and bought back part of the short side. The gross leverage and risk were kept under control, but some portion of the value was lost. This is something that we never – ever – want to do again. For that reason, we will slowly reduce the Fund’s gross leverage to a level such that even a market collapse similar to the one that occurred in October 2008 would not force us to sell anything.
It is quite possible that markets will have heightened volatility in the foreseeable future. In that environment and if we continued to maintain the gross leverage at 3, we would have to add a third condition for the success of our investment strategy – and that is that our stocks would need to outperform markets at all times. That would of course be too difficult, if not impossible.
We have already begun the leverage reduction and have a medium-term target for gross leverage of 2. This means that when fully invested the short side would be about 50% of capital and the long side about 150% of capital. That combination gives the gross leverage of 2. The reduction will be fairly slow in order to maximise the portfolio potential. With gross leverage of 2, the Fund’s investment strategy will be more conservative but with still significant growth potential.
Being more conservative in this case does not necessarily mean lower future returns. When we have selected a stock into our portfolio, we not only considered whether it is a good investment but also if it can beat the market. We have always needed to beat our short side. This perhaps led us over time to more aggressive investments than would be optimal. It forced us to work harder and come up with really good investments, but in the market crash it was exactly these investments that performed the worst. It was in these stocks that we saw the most forced liquidations from other funds. That was quite understandable, because we could hardly have been the only ones that saw great potential in these investments. They attracted the most speculative capital and were punished the most later. We continue to hold most of those titles, and they will probably drive our future returns. We will still be looking for similarly attractive investments in the future, but we perhaps will not be quite so aggressive in our stock picking.
The selection of individual stocks is in normal times the most decisive factor in concentrated portfolios like that of Vltava Fund. It was relatively less important in 2008. We have made some great investments and also some that we wish we had never made, but that is always the case and will likely remain so in the future.
It is one of the paradoxes of life that sometimes when you work the hardest, the results are the worst. That was our 2008. We have tried to learn as much as we could from the experience and will endeavour to make the journey to the point where we all make money again in Vltava Fund as short as reasonably possible.
Daniel Gladiš February 2009