Portfolio reporting: IRR relative returns

London holdings up to date

I’ve belatedly updated the Obliquity London portfolio page in the same style as the other (holdings, reweightings, disposals tables) with up to date data, which shows the full history of the portfiolio. Most of the slow movements have been reported in blog posts but the page wasn’t up to date pending an ever procrastinated automation. No automated process yet but this is backed by a little toolset I’ve been building in numeric Python (pandas) which should make future reporting easier.

IRR benchmark units reporting

The concept I’m playing with at the moment is using IRR (Internal Rate of Return) on cash flows expressed in benchmark units. That is for any given position I extract all the cash flows (trades, dividends and corporate actions), which I get from my broker in sterling, and then convert them in accumulating units of a benchmark ETF, as if it was a currency — remember that in essence any tradable paper is a currency. So each sterling flow is translated to an ETF units flow at the price (“exchange rate”) on the day of the cash flow. This shows the balance someone doing nominal accounting in benchmark units, or equivalently using the benchmark as their “cash” asset, would see.

This makes relative performance very clear: the sign of the nominal PNL (portfolio value change) tellls you if you have out- or underperformed the index. Basically if you had funded every buy by selling units of the benchmark, and bought them back on sales, it tells you if you’d have more or less units following trading than passively sitting on the benchmark units.

The use of the IRR also implies a normalisation of time effects, to capture that a N% change over 2 years is not the same as N% over 2 months, which is hard to see in classical nominal PnL reporting.

Stamps IRR charts

So here are a few of the result for the small cap London Stamps portfolio, valued as of 2015-05-08. This is early software so subject to errors and bugs, the numbers have merely passed a plausibility test.

The benchmark is iShares MSCI UK Small caps (CUKS) which is a good substitute economically (I’d happily buy it as a replacement allocation if I stopped playing stock picker in this segment) though a relatively poor short term benchmark technically. The reasons for that is that it uses a worldwide definition of small caps, which basically in the UK market captures the bottom of the FTSE250 — which are traditionally viewed as midcaps in the UK markets — and the top end of the local small caps section. So the average market capitalisation is significantly higher than my mostly AIM oddballs stock picks. A pure AIM index index wouldn’t be a good benchmark either as it would have a would bunch of junior resources stocks and overseas scams that I don’t touch, and I would never buy an AIM index as an economic substitute. The MSCI methodology is pretty good at excluding the darker corners of the market, so it is a valid benchmark in sector and industry terms.

Predictably the portfolio as a whole has underperformed by almost 10% since inception in money weighted terms. We’ll blame midcaps doing well while really small caps had a lacklustre year and say the jury is still out on my stamp collector skills or lack thereof.

Now, let’s do some digging down. Here is the IRR of each stock since inception, including closed positions, as a monthly return in benchmark units.

London Stamps IRR Ranks 2015-05-08

This is not very precise chart because the more recent positions produce outlier IRRs — the computation makes less sense in the very short terms. This excludes positions younger than 100 days (only AMD at the moment).

Other than that no surprises really, and apart from the GW Pharma pot bubble the next winners are tech companies that got taken over (indeed I held them for relatively short periods).

Now let’s watch our underperformance by plotting an histogram of these same returns (with AMD back in).

London Stamps IRR histogram 2015-05-08

So we’ve got the reasonable level of symmetry, as predicted by theory, the problem is only that zero is on the wrong side of the chart for the time being. It’ll shift if the performance reverses.

A thing to note in these charts is that the average monthly moves are not big, mostly within the -5/+5% range, which implies that with the typical spreads in small and micro-caps, long holding periods are essential. This is well known but nice to have another confirmation. Anyone flipping their portfolio every couple of months, in addition to having trouble to find alpha, would get trounced by transaction costs. It should also mean that long term returns should be better, so let’s do a little trend check by plotting our returns against holding period (remember this includes closed positions):

London Stamps IRR Age plot 2015-05-08

That one looks good. Patience pays so far.

That’s all for IRR analysis for today. Let’s finish with a little check we don’t have any undue overweight, with a simple visual check of portfolio weights for current holdings:

London Stamps weights 2015-05-08

So far so good.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s