What I’ve learned as a Portfolio Manager

Thoughts on handling markets and other notes

Memo to: Vega Capital clients
From: Scott Shuttleworth


Neither the information nor any opinion, contained on this Website constitutes a solicitation or offer by Vega Capital Pty Ltd (“Vega Capital”) or its affiliates to buy or sell any securities, futures, options or other financial instruments. Decisions based on information contained on this Website are the sole responsibility of the visitor whom we would encourage to seek relevant advice from a licenced professional.

 

For those readers who know me, you’ll know I’m very young on the scale in terms of being a Portfolio Manager. I do read widely and take my craft very seriously and it’s been interesting to see the mathematical theory of investment management play out in the real world.

In today’s blog, I want to highlight some of the more philosophical lessons which have stuck out for over this year in running the Vega portfolio.

 

1. Recognise bias, here’s just one.

The biggest way in which bias shows up (in my view) is commonly known as ‘buying high and selling low’. Investors do this all the time.

Say a stock begins to rocket up. It then gets bought more as investors come to believe that it will keep on going. Many investors will even take more notice of bullish research than that of bearish research regardless of what’s going on with the fundamentals, simply to justify why they should buy.

Likewise, as the stock price falls, investors begin to take bearish research more seriously. Yet all that has changed is the price!

Take for example recent moves in the S&P 500. When the index hit 2,600 back in October (down from 2,900), the news changed from bullish to discussing the possibility of a recession.

Sure, there’s more scrutiny over trade issues and rising rates, but the fundamental backdrop is just the same now as it was just a few months ago. But investors repeatedly, despite knowing that it will cost them in the long run, continue to be biased in loving what’s worked recently and fearing what hasn’t.

The index falling by 10 per cent shouldn’t influence the probability of the economy going into a recession. Similarly, price movements shouldn’t influence an investor as to what investment thesis they believe in. The base case may change depending on the fundamentals, but certainly not price.

 

2. Recognise the influence of politics.

Politics within an investment team can create the same issues which price bias does – creates favouritism for one thesis over another independent of the information at hand. Put another way, politics in the investment process is the inverse of returns.

As many would know, I’ve recently begun interviewing candidates for job opportunities here at Vega Capital. My greatest focus in this process has been to try and develop a team which can tackle investment questions in an academic and balanced manner in order to ensure that this bias and any other biases will not influence our decision making.

 

3. Develop systemic decision processes based on objective measures

In order to combat our human wiring, investors should develop rigorously tested systemic decision trees which incorporate the information relevant for making investment decisions in a consistent way and rational way.

This idea is useful because regardless of what the market is doing, we can rely on our investment process to do the right things to generate returns over the long haul. Over the short run, things can be more variable, but since we’re not day traders, we have this patient luxury.

Such an approach also allows for backtesting so we can get a sense of how valid our theories are and how stable they appear over time. Now yes, many would respond ‘I’ve never seen a bad backtest and hence its value is limited’. But that’s a very weak argument. Firstly, why would you see a bad backtest? If it doesn’t work, why would it get past the development stage? The point of the backtest was to find evidence to reject a hypothesis after all.

And when it comes to a backtest which looks good, questions should then be asked about its repeatability and sustainability.

That said not all systemic process are equal. For example, if your model is based on correlations…well, that’s not really what I’m talking about I’d highly advise against building models from correlations unless you really know what you’re doing.

I know that as a quant in markets, correlations are supposedly my bread and butter but as a rule I actually use them very little in the Vega process because they tend to be so unstable. Anyone remember the RBA’s research on US stock-bond correlations from a few years back. The below chart summarises their findings.

What once had a negative correlation ended up with a positive correlation and if you’re running a portfolio based on this analysis you likely won’t know it (until it’s too late).

 

4. Remain humble, remember that you know so little about forecasting the future but always buy value.

None of us know for sure where the market is about go to, for example, it was just a few weeks ago when the S&P 500 hit 2,800 and last week when it fell to 2,620. Those at 2,800 believed the correction was over and we were headed for 3,000 and yet we're now back to 2,700 – Vega notably took a different view and started buying hedges, but not because it could forecast the moves, just that the hedges became too cheap.

As human investors, our emotions can’t help but take the last move and project it forward, even when the statistics tell us that markets mean revert more than they follow through (this has been reinforced in recent years by algorithms which trade counter trend at 2 and 3 standard deviations).

Our saving graces are to buy what’s cheap, sell what’s expensive, take a long-term view and only adjust our views for big debt crises (as an aside, I recently read Ray Dalio’s new book on crises which I’d highly recommend).

But even then, our emotions may begin to doubt a well-analysed view unless it becomes validated almost immediately by price action - and this is irrational. If we find something which say is trading for $48 but we think is worth $70, it may not go up to our valuation in a calm straight line but the thesis should remain intact - even if it falls to $40.

Volatility rules markets in the short term and if there’s any constant it’s that.

No Comments Yet.

Leave a comment