My recent portfolio review methodology

Hello all,

With 70+ holdings, I thought I would whittle this down by 10 or so, just to improve the overall portfolio quality.

But how could I do this objectively, ignoring any bias for story stocks, or anchoring based on price paid or time held?

On my company post for HLMA, I referenced an ex-CEO’s speech. He mentions he would not purchase companies with a roce less than their target roce (which in turn, over time, obviously helped them achieve their targeted returns).

Also, I remember when a relative required hospital treatment, the hospital said they would not undertake an operation unless they thought the chance of success was greater than their current average success rate, thereby increasing their ‘stats’ over time (sad but true!).

So, combining the above with:

  1. Greenblatt’s methodology (allocating a rank for each metric, for each individual portfolio company, and summing the rank for each company by metric), the metrics being:

  2. Fundsmith’s preferred metrics - roce, gross margin, operating margin and cash conversion (I ignored debt because non of my holdings have significant debt) ,

I got to an ‘objective’ system by creating my own ‘Frankenstein’s monster’ :japanese_goblin: :japanese_ogre: ranking system. I eliminated/sold my least favoured 10 holdings from the ‘below average’ section (bottom half) of the ranking system, and re-invested the proceeds within the ‘above average’ section of the ranking system, in my most favoured companies (which were already held).

This in turn increased the average ‘look-through’ metrics of my portfolio overall.

(It depends what sort of an investor you are, whether the above is useful to you. I tend to ignore forecasts, stories and turnarounds, instead looking for a trend of increasing chosen metrics).

In the future I intend to use the above to look for:

  1. drops in ‘rankings’ for each individual , when compared to the previous quarter, as an initial source of investigation, as well as

  2. screening for ‘swaps’ - i.e. looking for listed companies with metrics far above current average portfolio metrics, and ‘swapping’ them with below average portfolio companies. This should increase my portfolio average metrics further upwards over time.

Further, I think I am just going try to trade quarterly, to ignore market ‘noise’.





Thanks that is really interesting.

I like the idea of a Terry Smith-like portfolio look through. Is it relatively easy to do on Sharescope?

I also think systems or principles are helpful both to aid invidual stock decision making and to force relative judgements accross the portfolio. As I am sure you are aware, Richard Beddard has run his decision engine for more than a decade.

Greenblatt himself has returned to provide retail funds, under the Gotham brand, in the US. I believe these are variants of his magic formula (ranked EBITA and ROCE) with some hedges to reduce volatility for the retail investor.

I have less than 10 stocks (most of my investments are in funds or ETFs) mainly in micro growth companies with some value/special sits, so the ranking is probably less helpful.



Glad you found it interesting!

It’s easyish on Sharepad to get to lookthrough assuming each company has equal weightings. I’m unsure on Sharescope but it should be similar.

On the blue table layout, perhaps create a new layout with the metrics mentioned, or whichever metrics you are interested in.

Select your portfolio.

Right click on the column header/title and select ‘column statistics’. It should then show the average value for that particular portfolio and metric (but for a portfolio of equal weightings).

(You would need to download individual company values and weight accordingly in a spreadsheet if there are materially different value holdings within your portfolio).

Hope that helps.


Hi anon,

Sounds a bit like the approach taken by Stockopedia’s NAPS portfolio (subscription required), which has worked very well over time. Objectively looks only at the numbers and ignores all the human biases that hamper us investors.

Would be interesting to know which 10 shares you dropped and which shares received the proceeds :slight_smile:



Yes, of course! - Be warned, some are US shares :us::


US :us: - FSI, FIX, AMZN, ACN, ZBRA, EPAM (AMZN because unsure of their retail business and lack of synergy with Amazon Web Services - ref Fundsmith AGM comments this year).

UK :uk:- FOUR, THS, EVR, UPGS (Unsure of mining as an investment proposition at all any more, and worried about companies reliant on goods shipped from China).

These selections were all from the bottom half of my ranking system.



UK :uk:- JDG, SAFE

(EL :us:and NKE :us: are actually new positions, but their metrics are well above my average portfolio metrics - I just didn’t want to complicate my initial post unnecessarily).

If these were 2 separate portfolios I think I would definitely rather own the ‘bought’ one than the ‘sold’ one, so I am happy with that.

You may disagree, though?



1 Like