In an earlier post, I described my approach to benchmarking for individual investors. It’s centred on three main ideas:
- Comparisons matter. Chosen benchmarks must facilitate “apples to apples” comparisons.
- Creating meaningful feedback. Investors need to dig a little deeper and create feedback at the decision-level, not just the overall portfolio level.
- Chunking. break your performance down into “chunks” that represent the key investment decisions that you make.
The table below summarises the main “chunks” that I use when benchmarking my personal portfolio. These chunks work for me. Ideally each investor will develop their own “chunks” that align with their investment process and time horizon.
I have been running a small personal portfolio of US stocks for a little over twelve months. The portfolio currently contains 6 stocks. (I aim to keep my portfolios between 6-15 stocks).
I would be lying if I said that I didn’t look at my performance daily. It’s a bad habit and I’m trying to quit. But I only formally assess the performance of my portfolio once every 12 months. This is when I break down my performance into the five criteria or “chunks” to identify what I can improve.
The following charts and comments are a small sample taken from this exercise. Hopefully they will give individual investors some ideas on how to benchmark their performance.
The charts below are taken from Google Finance. They show the dollar-weighted profit and loss (not the total portfolio value) of my portfolio (top panel) and a notional portfolio featuring the SPDR® S&P 500® ETF (SPY). Values are in USD.
Two things stand out. Firstly, I under-performed the SPY by approximately $2,000. Secondly, my portfolio returns are more volatile than the SPY ETF.
Am I bothered by this result? No, not really. It is no surprise to see that a 6 stock portfolio (which only became a 6 stock portfolio 9 months into the year) is more volatile than an ETF tracking the S&P 500.
The whole point of owning a 6-stock portfolio is for it to perform differently to the market (see portfolio construction). Naturally this is a two-edged sword. Sometimes “different” will mean under-performing the market over the short-to-medium term.
Even the best fund managers routinely get 40-60% of their decisions wrong. Take the Tiger cubs, some of the most well-known hedge funds, as an example. According to NOVUS, their batting average is 56.92% (but who cares with a 2.6 to 1 win/loss ratio!).
12 months is too short a time frame to assess my skill level. It doesn’t allow enough time to learn from my mistakes and improve as I go along.
Importantly, 12 months isn’t enough time for me to become desensitised to making mistakes. Every investor needs to learn for themselves, by experience, that it’s the performance of the portfolio that counts. This mindset doesn’t come naturally, especially to perfectionists with a competitive streak such as yours truly.
I’ve made some changes to my investment strategy that have improved my performance. These changes a as direct result of feedback that I’ve created. They have resulted in improved performance so this year.
Its impossible to learn how to manage your behaviour by reading a book. The only way is to get out there and do it! I’m hopeful that I’ll continue to refine my investment strategy and improve my results. I think of my under-performance as the”tuition fee” for learning how to become a better investor.
But if I significantly under-perform over a period of 3-5 years then I should think seriously about sacking myself and investing in a passive index fund.
My under-performance was primarily due to two reasons. Higher than normal trading costs explain almost a third of the under-performance (see Efficiency).
The rest was due to two small speculative positions and a core position in which I averaged down (increased my position as the price fell). These positions lost -17.14%, -21.00% and -23.98% respectively (see selection and portfolio construction).
There were 13 trades (11 buys and 2 sells) in my portfolio. This is more than I would ordinarily trade since I started with 100% cash on day one and had to purchase stock to construct my portfolio.
My notional SPY portfolio had only 8 trades (8 buys). The brokerage paid on the additional 5 trades explains approximately 28.2% of the under-performance between my portfolio and the notional portfolio.
The lesson that I took from this feedback is this: unless I have an edge, each time I trade, I shift the odds in favour of the passive ETF benchmark.
Tax wasn’t really an issue as the two sales were both at a loss.
I expect that this year’s brokerage costs will be lower due to starting with an established portfolio. However, the feedback on my trading costs has prompted me to review my brokerage arrangements which will hopefully make a big difference.
Selection refers to checking if the stocks that I invest in are consistent with the types of opportunities that my investment strategy is designed to identify. I learned a few things by going over my investment strategy. For example, last year I carved my portfolio up into two buckets:
- Good companies at a fair price
- Special situations – distressed companies, turnarounds, deep value, etc.
The detractors to my performance came mainly from the special situations category. I discovered that I don’t really have an edge investing in these companies because:
- The lower hit rate means that comes with this sort of investing means that a more diversified (at least 30 stocks) portfolio is appropriate.
- Splitting my portfolio up into smaller lot sizes (which I would need to do to manage risk) would significantly increase my transaction costs.
- Portfolio turnover is higher which increase brokerage costs and taxes (if I’m successful).
- Special situations require more a lot more work and ongoing monitoring.
- I simply don’t have the time to analyse and cover more than a handful of stocks at the moment.
My portfolio is currently mostly made up of good companies at a fair price and a large allocation to cash.
The lesson here is the importance of honest self-examination. It’s the only way to learn the boundaries of your circle of competence.
I use daily closing prices of the stocks in my portfolio and the S&P 500 to calculate the following summary statistics for my portfolio.
As you can see, my portfolio was riskier than the market (higher beta and volatility). Performance was also less consistent: 13 more down days and 12 fewer up days. This doesn’t tell me much other than I need to keep digging to understand why.
The next chart shows the evolution of my portfolio allocations over time. One of the reasons for the higher volatility/lower consistency of returns was that I took a few months to diversify across 6 stocks. Two of my early purchases were the small speculative positions that later exited at a loss.
Here is the composition of my portfolio at the end of 12 months. By this time the portfolio is more diversified. Equal weighting is a good default but I’m happy to have a larger position in a stock where I have high conviction. For example, the stock shown in orange is almost 30% of my portfolio.
I dug further to better understand the higher volatility/lower consistency of returns by looking at the daily return contribution of each stock to the overall portfolio.
This yielded some interesting insights:
- Stock H (brown) generated a lot of volatility, even though it was a small position.
- Stock F (green) shows two large downward spikes. Here the stock price was reacting to disappointing quarterly results.
- Stock E (light blue) generated a lot of volatility in the early part of the year because it was one of only 3 stocks held as I was building the portfolio.
The chart below repeats this analysis over a 20-day (roughly a month of trading days) horizon. This highlight just how much of the variability of results was generated by stocks H (brown) and F (green).
Stocks F and H were the two small speculative positions that I sold out of during the year. After that the variability of my results reduced.
The analysis illustrated above is a good visual check on my portfolio construction. I didn’t realise just how much my smallest and speculative (i.e. risky) ideas were driving performance and volatility. That’s not good when running a concentrated portfolio.
I also track each individual buy, sell and hold decision. I do this in three ways:
- Keeping track of the absolute number of buy and sell decisions that I get right (batting average or win/loss ratio). I keep a tally both in absolute (loss vs gain) and relative (vs S&P 500).
- My dollar win/loss ratio. In other words, the percentage of my portfolio allocated to winning vs losing stocks. Once again, I track this in absolute (loss vs gain) and relative (vs S&P 500).
- Investigating the impact of my behaviour (buying, selling and holding decisions) when managing each individual position.
The table below shows the win/loss and dollar win/loss ratios for all of my buy decisions.
As you can see:
- Only 45.5% of my buy decisions or stock picks resulted in an investment return greater than zero and only 36.4% beat the S&P 500.
- My dollar win/loss ratio was much better. 57.9% (win/loss ratio = 1.37) of the money that I invested earned a return greater than zero and 45.6% (win/loss ratio = 1.20) beat the market.
How do I interpret this? My picking was bad but my ability to size positions was good as I invested more money in the best performing stocks.
Once again, the impact of the two speculative stocks can be seen. Had I not invested in these stocks, my batting average and win/loss ratio would look like this:
- 62.5% of my buy decisions or stock picks resulted in an investment return greater than zero and 50% beat the S&P 500.
- My dollar win/loss ratio was much better. 68.5% (win/loss ratio = 2.17) of the money that I invested earned a return greater than zero and 53.9% (win/loss ratio = 1.17) beat the market.
This reinforces the conclusion that it’s too soon to draw any firm conclusions. A handful of investment decisions is too small a sample size.
I create charts to analyse my trading behaviour in response to changes in price, absolute and relative performance. Here’s how to interpret the charts:
- Dark blue line (left axis) shows the cumulative holding performance (price only) of the stock over the year. A dark blue line above (below) zero means the absolute performance of the stock is positive (negative).
- Dotted black line (left axis) is the cumulative performance (price only) of the S&P 500. The stock is out(under)-performing the S&P 500 if the dark blue line is above (below) this line.
- Grey shaded area (left axis) represents performance spread between the stock and the S&P 500. The stock is out(under)-performing the S&P 500 if the are is above (below) zero.
- Yellow dashed line (right axis) shows the number of shares held.
You could also create similar charts examining other variables, for example price-to-earnings ratios. Whatever helps you to analyse the decisions that you’re making.
The chart above is for Stock A, my best performing position. I initiated my position in Stock A in May and held it through the rest of the year. The stock was very cheap relative to its fundamentals at the time of purchase. In fact, it continues to look cheap relative to the market, especially if you strip out its large cash balance.
The question that I take away from this chart is this: Why didn’t I buy more around December 2016/January 2017?
I had identified this stock as a core long-term holding (i.e. not a speculative or relative value opportunity), the outlook was improving, the fundamentals remained solid, market momentum was positive and the valuation was still cheap (although not as cheap as before). So why didn’t I buy more? My guess is that the disposition effect got the better of me. Like many people, I find it very hard to buy more of a stock once it goes up. It just felt wrong, even though it was the right thing to do.
So what’s the lesson? Peter Lynch’s was right.
Peter’s Principle #11
The best stock to buy may be the one you already own.
Many stocks which later became major holdings started out as minor purchases by Magellan. Often it is unnecessary to run around looking for the perfect stock, you may already have it in your portfolio, so buy more! It goes back to principle number 9, the number of really brilliant companies is finite, so when you do have one it might be better to buy more than to go out to find something else.
I now have a rule where I don’t add a stock to my portfolio without first checking to see if I should be adding to an existing holding. Analysing my performance has given me a vivid example of why this is important. Hopefully this gives me the nudge that I need to buy more of a stock if it’s the right thing to do, even if the price is now higher.
Losses often result in the best lessons. Here’s the chart for Stock H. Stock H was a small speculative holding. The company was a serial acquirer that had recently been embroiled in several scandals. It was also highly-indebted. Stock H had suffered a precipitous decline. The largest investor had finally capitulated and sold out for a massive loss.
Why would I buy such a diseased company? I don’t have a problem with allocating a small part of my portfolio to turnarounds and other speculative opportunities; provided that I go into them with my eyes open. My opinion was always that Stock H was a long shot. It had a high probability of going to zero and a small probability of increasing 2-3x in value in a short time. So I made sure to:
- keep my position small
- pre-committed to a definite stop loss
I assumed that most of the bad news was probably reflected in the price. Nearly all of the management team and the board had been replaced. Some of the acquired businesses were profitable and there was a plan in place to reduce debt by selling several business units. If they could pull it off, their might be upside. So I bought a small position in August. The market reacted favourably to several announcements by the company, so I increased my position in September.
Then things took a turn for the worse. Despite assurances from management, Stock H’s scandalous behaviour continued. The company struck a deal with creditors to capitalise interest, further increasing its debt. The asset sales that were supposed to reduce debt now looked less certain. Investors were also voting with their feet and the stock price reversed and was now falling just as quickly as it had bounced.
I sold out in November, as the news surrounding the company deteriorated and the stock price was nearing my stop. I booked a 17.4% loss, but I avoided the almost 70% loss I would have suffered if I continued to hold the stock!
I am more proud of my behaviour with Stock H than I am of Stock A. Why? Because selling as a loss is hard to do (disposition effect and regret aversion) and because I stuck to my investment process. Stock H is a visceral reminder of the importance of being honest with myself (am I investing or speculating?) and risk management.
Yes, you could argue that I probably shouldn’t have invested in a long-shot (and you’d be right). But at least I protected myself by using a clearly defined risk management strategy and I dealt with my mistake quickly before it blew up my portfolio.
I have a lot of areas to improve on. More importantly, I have a clear understanding of what those areas are and a plan to work on them. I couldn’t have done this without the feedback that comes from creating meaningful comparisons.
So don’t just compare your performance to a benchmark, you need to make one that you can learn from!