Wisdom From Terry Smith + The Most Important Investing Ratio*

23rd Jan by Roland Head

31 comments 7511 reads

There’s no doubt that Warren Buffett’s annual letters are a worthwhile read for any keen private investor. But personally, one of my favourite missives comes from a fund manager who operates on this side of the Atlantic – albeit with a penchant for holding US stocks.

Terry Smith’s 2023 letter was published a few days ago and offers the usual mix of pithy opinion, investing insight and Fundsmith performance metrics.

Starting with the latter, Fundsmith’s flagship Equity Fund gained 12.4% last year, underperforming its benchmark the MSCI World Index, which returned 16.8%.

I don’t think this is too bad a result, for a year in which a big chunk of global equity returns were driven by the so-called Magnificent Seven, of which Smith owns some but not all. (The Magnificent Seven are Alphabet (NSQ:GOOGL), Amazon.com (NSQ:AMZN), Apple (NSQ:AAPL), Meta Platforms (NSQ:META), Microsoft (NSQ:MSFT), NVIDIA (NSQ:NVDA) and Tesla (NSQ:TSLA))

Importantly, Fundsmith remains ahead on a long-term view, with an annualised return of 15.3% since inception in 2010, versus 11.5% for the MSCI World Index.

Smith’s secret sauce (and my top metric)

I think that some of Fundsmith’s outperformance has been driven by significant multiple expansion over the last decade or so, in addition to good stock selection.

But I’m in no doubt that much of his success is due to his disciplined focus on an investing ratio I would bill as the most important metric in long-term investing – return on capital employed.

Smith never fails to emphasise this in his writing and sure enough, ROCE gets a mention in this year’s newsletter:

Consistently high returns on capital are one sign we look for when seeking companies to invest in.

There is a caveat, of course:

Another is a source of growth — high returns are not much use if the business is not able to grow and deploy more capital at these high rates.

I think the significance of these two requirements and their real-world impact can best be illustrated with an example.

Compare the share price performance of FTSE 100 members Tesco (LON:TSCO) and Unilever (LON:ULVR) over the last 20 years:

spUQ5sFouINKLdOr3evjpZT1Yiao_MkRGc-4799GhJtBnUH1PLx7TyNesYnYWrC09VOvgjCiH3sZvd40Z8c31kDORPieqpY-4bOs9U9PsME0DIo92S4Di_XStY1ftAtYnNWyGb0EijSDtEvgN1sf1yk

In general terms,…

Unlock with Email

Unlock with Google

Already have an account?
Login here

Indie

23rd Jan

Hi Roland, great article - always helpful to have the fundamentals explained so clearly.

When I go to add ROCE to a screen, Stockopedia has 5 different ones:

TTM
Last Year
5y Avg
Long Term Avg
Prior TTM.

I see that you use Long Term Avg in your screen. Am I right in thinking that the calculation you do in the article for James Halstead would give you the Last Year ROCE? When would the different ROCEs be useful?

DWit199

My figures are total return including dividends. The universe is UK, member of FT All Share, Fledgling or AIM All Share index and mktcap >£20. It excludes the International Order Book and a few other oddities that Stocko seem to include. A few shares will be accidentally omitted and there is an element of survivorship bias. Within those limits, I am reasonably confident that the data I have presented is correct and representative of my UK investing universe but it probably wouldn't pass peer review for an academic paper. You are welcome to review the whole data file to check its integrity and repeat the analysis if you wish.

Edited to add that I have uploaded the file to WeTransfer. It can be downloaded from this link
https://we.tl/t-ezGy6NByg5

Rusty2

I'm not suggesting your figures are wrong but there is a fair difference, some is the dividends which Stocko does not include. Also they have all stocks over £10m mcap. Also the rankings are different, I was looking at 80-100 ranked stocks.

KevinS1

I use a number of ETF's that have the "quality" factor built into their method - these normally look for a high ROCE, low debt and a consistently low YoY earnings variability.

I tend to use these benchmarked alongside the Fundsmith fund, which has an excellent long-term record, but it is high-cost and its short-term record is not that great.

Only time will tell if they are any good - Wisdom Tree has a suite of them, but they are not the only ones.

SCI20

Are you referring to U.K. specific ETFs?

CautiousValue

I guess that's the hard part, finding those companies that can keep their ROCE high for ten+ years. Are you able to see data on how persistent ROCE is? e.g. if a company ROCE was >20% between 2000 and 2010, how likely was it that remained true for the next ten years? Or even does an improving ROCE drive the share price to outperform?

Guinness Global Equity Income run their entire fund on a similar basis, using the CS HOLT system.

Roland Head

Hi Indie,

Good question! The James Halstead (LON:JHD) example is for the last year's ROCE. At the moment, last year will be the same as TTM (trailing 12-month) for Halstead, because its most recent accounts were full-year figures.

The TTM figure is useful for a more up-to-date figure after a company has published its interim results - in this scenario, the Last Year figure can be quite stale. When applicable, the TTM figure is shown in the Quality box on the right-hand side of the StockReport and in the main financial tables.

My use of the long-term average ROCE is to smooth out any short-term fluctuations. For example, if a company reports a statutory loss one year then its ROCE will appear negative. But for a good quality business, this is hopefully a temporary glitch and not representative of its normal performance.

Hope this helps,

Roland

DWit199,

Over a short period such as one year, I would also expect the StockRanks to be a more likely predictor of performance, due to the StockRank's more mix of factors, including momentum and valuation.

In my view, ROCE may be a more useful indicator for (very) long-term performance, when the impact of reinvested earnings and compounding can become much more powerful than short-term valuation changes or momentum.

No - I have UK, US & Global quality ETF's; WisdomTree Global Qual Div Gro UCITS ETF USD Acc (LON:GGRG) , WisdomTree UK Quality Div Growth UCITS ETF GBP (LON:UGRW) , UBS Factor MSCI USA Qlty ESG UCITS ETF USD Adis (LON:UC99)

Wisdomtrees website has a specific link in their menu which points you to their whole global Quality factor offering - they have four, with specific options in each one for currency & hedging options.

Sunil

Fidelity (FGQD) and Blackrock (WQDS) have Quality factor ETF's too, both with decent liquidity.

Neil McQuillan

24th Jan

This article is excellent, thank you. A really good read, the standard of discussion in the comments is great as well. Makes my subscription worthwhile.

My advice to anyone looking at ETF's would be to look at them using the Just-ETF website where you can compare them alongside one another and determine which one best suits your needs - holdings (both specific & total number), costs, size, dividends, volatility, drawdown etc. So many different things to look at and to find your own individual requirements.

I then also tend to benchmark them against other IT's and UT's to see if it is worth paying the extra fees for outstanding performance.

There are also some good videos on Ian Shadracks you-tube channel looking at a lot of funds/ETF's/IT's.

Good luck.

Thanks, Roland, very clear and comprehensive - as usual. At the moment, my table view has TTM but I think I'll add 5y Avg and maybe Long Term to get used to the differences with a view to using a longer-term ROCE for screening, now that I've been reminded they exist.

Yours are the articles that I come back to and re-read the most.

Epitychia

A fair point, ROCE does decline in the long run. The key is to find those with a wide enough moat. Buffett is good at this, but for a mere mortal it is hard.

if a company ROCE was >20% between 2000 and 2010, how likely was it that remained true for the next ten years? Or even does an improving ROCE drive the share price to outperform?

This is a very hard question to answer as the data is not readily available to a PI. There are so many corporate events that over a 10 year timescale the stats can be quite distorted by takeovers, mergers, delistings, name chages etc. With that caveat, I have looked at LSE listed companies where I have 3yr Average ROCE data from 10 yrs ago, current 3 yr average ROCE and total return data for 10 yrs.

There were 127 companies which had 3yr av ROCE >20% 10 years ago. Over that time the All Share Index had an annualised Total Return of 5.15%. Only 51 companies from that list had a TR exceeding that figure.

From the original 127 only 35 companies have a recent 3 yr average ROCE of >20% so the majority do not maintain substantially above average ROCE over the long term. Of this 35 a respectable 22 have managed to exceed the return of the All Share Index over the 10 year period so maintaining a high ROCE does increase the probability of long term outperformance.

The "ROCE 35" are listed in the table below so you can get a feel for the best of the best.

Please bear in mind these numbers come from a different data supplier so may not exactly match numbers available on Stockopedia.

TIDM	Name	10y ago ROCE 3y av.	ROCE 3y av.	ann % Total Return 10y
GAW	Games Workshop Group PLC	36	64	35.6
PLUS	Plus500 Ltd	313	85	24.9
BVXP	Bioventix PLC	46	76	24.8
CMCL	Caledonia Mining Corp	33	24	17.6
RNWH	Renew Holdings PLC	72	33	17.6
WIN	Wincanton PLC	25	23	13.9
DOTD	Dotdigital Group PLC	35	21	13.9
ASY	Andrews Sykes Group PLC	32	24	11.4
HWDN	Howden Joinery Group PLC	46	24	10.8
DOM	Domino's Pizza Group PLC	40	28	10.0
REC	Record PLC	30	38	9.9
CLIG	City of London Investment Group PLC	75	21	9.8
RMV	Rightmove PLC	316	214	9.2
CCT	Character Group (The) PLC	24	21	8.9
BHP	BHP Group Ltd	28	32	7.5
BLV	Belvoir Group PLC	1347	20	6.4
IGG	IG Group Holdings PLC	32	28	6.3
SPT	Spirent Communications PLC	26	22	6.3
PAF	Pan African Resources PLC	32	26	5.7
DNLM	Dunelm Group PLC	55	37	5.4
JHD	James Halstead PLC	42	30	5.3
NXT	Next PLC	58	25	5.2
FXPO	Ferrexpo PLC	35	41	5.1
PAGE	PageGroup PLC	40	28	4.5
WINK	M Winkworth PLC	48	36	3.8
POLR	Polar Capital Holdings PLC	23	36	3.8
JIM	Jarvis Securities PLC	114	117	3.0
STEM	SThree PLC	32	32	2.6
TEP	Telecom plus PLC	49	25	1.1
AEP	Anglo-Eastern Plantations PLC	21	20	1.0
PAY	PayPoint PLC	43	29	0.1
STVG	STV Group PLC	25	23	-0.9
HL.	Hargreaves Lansdown PLC	115	56	-2.7
SNWS	Smiths News PLC	210	75	-7.2
DLAR	De La Rue PLC	47	41	-12.2

NeilS

27th Jan

Howden Joinery (LON:HWDN) are building out in France now (pardon the unintentional pun!) and I think ROI.

ShoeShineBoy

28th Jan

Good article, thanks for posting Roland. Investing in companies with high ROCE and high margins turbo charged Warren Buffett's returns. It was Charlie Munger who convinced him of that as explained in the Joel Greenblatt book. Even so, I will keep hold of my Tesco (LON:TSCO) shares for the moment, they were a solid dividend payer when BoE rates were low and will be so again if BoE rates come down.

High quality podcast interview with Terry Smith recently. Well worth listening to.

http://tinyurl.com/yc5w3mkx

The 'problem' with Fundsmith is that it is now so large, investing in small caps doesn't turn the dial on bottom line performance so it can only really buy large caps. Smithson Investment Trust (LON:SSON) is a good alternative. I understand Terry Smith is not involved on a day to day basis but Simon Barnard follows a similar strategy to the open ended fund.

SSB

WDWombat

29th Jan

In principle I completely agree and am heavily influenced by Smith who succinclty wrote on this topic in the FT of March 16th/17th 2013. 'Sorting the Wheat from the Chaff'. But two caveats.

1) I am constantly wary of the use of a data comparison starting dates as they are so often used to distort a comparison or support an argument. It is true that over 20 and 10 years Unilever has basically clobbered Tesco. Over 5 years, however, the relative share performance is 40% in Tesco's favour.

2) In support of the article's recommendation to do your own maths, or largely accept Stockopedia's numbers, I am completely in agreement.

I recently bought Whitbread but had noticed, prior to purchase, that there was a huge discrepancy between the RoCE indicated in company releases (12%) and the same on the Stockreport here (6%). I queried this with management who replied, fairly, that a description of their calculation was in the back of the annual report as a note. They had excluded the business in Germany, which is very promising, growing but not yet profitable. The company has also been using adjusted rather than statutory operating profit, so (as far as I understand) has been using a figure closer to EBITDA than OP - that is pre-depreciation, pre-option awards. Or that is my understanding.

Timoteo

You have picked out the same glaring errors that I did, all of which were avoidable. I do wonder whether there is sometimes ‘institutional paralysis’ which hits fund managers because of the size of their holdings and the consequent difficulty in rebalancing. I can’t understand how a guy in Smith’s position ignored what was happening at PayPal for so long nor why he went into Tesco when he did. Unilever is even more baffling. Why hold such a large position in a company of whose management you are so critical.

I also think his marital situation may have impacted on his fund performance. I have noticed that people in many walks of life perform less well when they have a high score on the MDI (Marital Disharmony Index) !

Patrickmylnetrade

31st Jan

thanks Kevin great pointer.. More importantly huge thanks to Roland for a very interesting topic, and article

Wisdom from Terry Smith + the most important investing ratio*

Smith’s secret sauce (and my top metric)

Unlock the rest of this article with a 14 day trial

About the Author

Roland Head

31 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...