Christmas Commentary | 29/12/23 | DWHT/DWHA | Google is your friend | Screening for Bargains

This week, Jamie Ward briefly talks about the latest update from Dewhurst. He then demonstrates that markets are frequently not efficient by reviewing the performance of a couple of the COVID winners to see if it can inform future investment decisions. Finally, he shows how to use Sharepad to screen for businesses that attempt to view valuation on a through-the-cycle basis.

MCC Blog

With it being Christmas, not a lot is happening in markets, however, the release of an update from Dewhurst did send me down a little rabbit hole. I have an equity investing fetish. I like businesses that share a name with the surname of board members. That is to say, businesses that are still run by either the founder or, more likely, his descendants (I know, not very PC, but realistically, these are old businesses that were all founded by men; little point writing his/her). In the UK, these include:

  • Dewhurst, which basically makes buttons for lifts and trains as well as other environment interaction-type products and has two Dewhursts on the board.
  • Goodwin; a business that is well respected amongst the private investor community and virtually unknown by the professional investor community. It is an engineering business, although that rather sells it short because it does all sorts of things and is well worth looking at just for interest. Four of its board members are Goodwins.
  • James Latham, which is a supplier of specialist timber products and predates the American Revolution. Two of its board are Lathams and of goodness knows what generation to run the business. It also intersects with another interest of mine – that being woodwork.
  • James Halstead, which is a Manchester-based producer of flooring and is currently run by a fourth-generation Halstead with another one sitting on the board as an independent director.

The reason I like these businesses is that they are often run for capital preservation as much as capital appreciation and while that might sound boring, it means management are often focused on things that create very good outcomes for shareholders. These include strong balance sheets, opportunistic investments when bargains are available and a conservative appraisal of opportunity sets. Philosophically, this type of equity investment would fall into the category of ‘worry about the downside and let the upside take care of itself’. All of the businesses above have been listed for at least thirty years and despite doing rather old-fashioned(ish) activities in unglamorous industries, they have all trounced the stock market over the very long term. Their respective performances are as follows (all on total return basis with dividends reinvested for thirty years); Dewhurst, 2,189% for voting class; Goodwin, 44,707%; James Latham, 7,627%; and James Halstead, 3,179%. This is compared to 128% for the FTSE100, 224% for the FTSE250 and 226% for the FTSE Small Cap. The AIM market is not yet thirty years old but its 28-year return is dismal – down 25%. With businesses like those above available to buy, why would one purchase a tracker?

Of course, the company name doesn’t have to be the same as some of the board, there are more examples of businesses that are run by families where the business name is not eponymous:

  • AB Foods, known by most to be Primark plus some other stuff. In reality, it is a large diversified food business as well as a discount retailer called Primark. The family in question are the Anglo-Canadian Weston Family, who collectively own more than half of the business. It is currently run by George Weston, who I think is the fourth generation to run the business. The shares have returned 1,623% over the last thirty years.
  • Renishaw – this isn’t really a legacy of family holders type business but given the octogenarian founders own more than 50% of the equity and are still involved, it counts. The shares have returned 3,588% over the last thirty years.

Dewhurst, Preliminary Results for the year ended 30 September

A few weeks ago, I talked about how I disdained excessively ebullient language from management. I talked about it in the context of Pensionbee, which was rather over the top in the exciting language used to describe a business transitioning from losing money to, erm, losing less money. With Dewhurst, we saw the antithesis to this sort of roseate phraseology in its update from December 21st despite the business generally doing OK.

Full-year results

It wasn’t exactly a vintage year for the company. As is so often the case, the parts of the business did well but were offset by disappointment elsewhere. In the main, revenue was up very slightly, though one assumes that volumes were probably down a little given the inflation of the last twelve months coupled with the meagre increase in sales. Last year, the company suffered a cyber-attack, which cost £1.5m to remediate. The effect is that, whilst adjusted profit is down slightly on account of lacklustre sales, the reported number is higher given last year’s drag.

Overall, the business delivered a modest increase in per-share earnings on a reported basis although I would surmise it was down c. 13% on an underlying basis. With these sorts of businesses, where management are so heavily invested, there is often a very conservative approach towards both capital allocation and disbursement. In the case of Dewhurst, this means that the dividend is roughly one-quarter of the earnings, whilst the low levels of growth mean that capital expenditure requirements are also very low. Taken together, despite the lacklustre performance, management were able to propose a 7% increase in the total dividend for the year. On a through-the-cycle basis, this 7% is probably what management think roughly that the business can sustainably grow at and indeed, would be in the ballpark of the long-term achieved levels.

Given the low levels of capital expenditure and small dividend, it probably wouldn’t come as a shock to discover the absurdly strong balance sheet. Cash on the balance sheet now sits at over £24m, an increase of £2.5m on the prior year. To give context, the market capitalisation is £52m and the business generates roughly £5m per year in net income. There is a small pension deficit but this is dwarfed by the cash pile.

Outlook

Compared to the aforementioned Pensionbee, Dewhurst sounds rather dull – they often do in fact but, despite this, its long-term performance is admirable. There was a destocking issue in one of the major customers but that appears to have abated. The bout of inflation hasn’t helped as some contracts were fixed price, but management seem happy that the ability to respond to further inflation has improved. Management also believe that the cash pile affords a great deal of flexibility to deploy capital should opportunities arise.

This last point, ultimately to me, is why I like conservatively run businesses. Some may believe that the balance sheet is inefficient and the cash should be reduced via a special dividend or the like, but a bit of dry powder never hurt anyone. Also, given where rates are now, that cash will be generating a meaningful return by itself.

Opinion

Dewhurst has struggled since COVID, but that period of difficulty looks to be coming to an end now. Considering the value created and the probably 5 to 7% growth rate of the business, it looks very cheap. It is a bit of an oddity in the UK as it is one of only two companies that I can think of with voting and non-voting share classes – the other one being Schroders. For a long time, the non-voters traded at a ridiculous discount to the voters; at times approaching 60%. In the last few years, the large discount that the non-voters traded at afforded the shares a degree of protection. So that whilst the voters are down two-thirds in two years, the non-voters are down c. 30%. This also means that the discount is now only around 25%.

I like Dewhurst and have owned it for many years. Given where the discount sits today, were I to buy more, I would probably buy the voters. It really doesn’t matter which metric you use; the shares look very cheap to me. If you net off the cash position, the shares trade a little more than 7x earnings and those earnings could be at a low ebb.

Before demonstrating a method for how to filter shares for valuation in the context of the cycle, we will look at how markets, however rational in the long run, prove to be far from rational or efficient at times.

Market Inefficiencies – Google is Your Friend

Following the release of figures from Tandem Group a few weeks ago, which appeared to be poor, I went down another rabbit hole. This time regarding illogical valuations in markets at times and potentially a way for private investors to gain an occasional edge.

During COVID, Tandem, along with certain other companies roofed it (technical term – it means “went up a lot”). The reason being, that suddenly a lot of people were flush with cash and nothing to spend it on like commuting, eating out and holidays. The result was a boom in spending on leisure activities, particularly those that promoted a kick-start to fitness.

In other words, people went out and bought bikes in their droves (or rather they sat at home and ordered bikes). I remember a story in the FT from about June 2020 that said a little independent bike shop in London was selling as many bikes per day as it used to sell in a month. Clearly, that wasn’t sustainable, albeit it would provide a meaningful fillip to the short-term revenues of beneficiary businesses. Although, a logical assumption would be that some of these revenues were brought forward from the future as people don’t tend to replace bikes often.

Nevertheless, the share prices of the companies involved repriced so high that they had to have been pricing in a permanent improvement in performance. I.e., the sale of bikes in what was a very warm and pleasant spring and summer coupled with the aforementioned consumer boredom wasn’t a flash in the pan but the new norm. It would take a special degree of credulousness to believe that. Below are the share price charts of Tandem and Halfords – the other obvious beneficiary of the boom in bicycle purchases.

In the case of Tandem, you can argue that its tiny size meant that any speculation in the shares could have pushed up the share price meaningfully and exiting a position becomes very difficult. However, in the case of Halford, no such argument can be made.

At this point, you might surmise words to the effect ‘well yes, in hindsight, we know that the share prices became silly but we didn’t know for certain at the time’. That brings us on to the bottom of my rabbit hole, what if there was a way of knowing contemporaneously that things were getting silly? To which I think I found the answer in Google Trends.

Google Trends is a free service offered by Alphabet that analyses the popularity of search terms and can do so with quite a bit of granularity. For example, you probably wouldn’t be shocked to read that the frequency of the search term ‘Christmas Gift’ over time seems to spike in late autumn and early winter but is entirely absent the rest of the year. The following shows the popularity of such a search term over the past five years.

Notwithstanding March 2021 when for some reason people started searching for ‘Christmas Gift’ for about a week, this is entirely predictable.

What if however, for some search terms, it wasn’t predictable but rather predictive of other things? In other words, what if these search trends were leading indicators to real economic activity and therefore provided an insight into the share prices of certain companies? Below is the result for the search term ‘Bicycle’. Note there is a seasonality because people buy more bikes in summer – it being a much more enjoyable exercise in the warmer months. However, in early months of COVID, there was a much greater spike than typical.

By January 2022, the search term had returned to its old typical levels that we used to see in the pre-COVID era – a return to normal. However, the share prices of those companies that benefited had not returned. The conclusion was that people were no longer buying anywhere near as many bikes but that fact had not been reflected in the share prices of the beneficiaries. What followed was a near halving in the share price of Halfords, however, were one to have used the ‘Bicycle’ search trend, one could have foreseen something missed by the market and consequently missed out on the pain.

I’m sure I’m not the first person to use Google Trends as a tool to gain insight into an industry and thus far I’ve only scratched the surface, but I’m sure there are data available to help private investors see something before the market and to then benefit.

The following is in no way a recommendation but, the house builders have been awful for the last couple of years, albeit they have perked up a little recently. However, is the ‘market’ aware that people searching for ‘New Home’ looks like it is recovering?

Finally, this week I demonstrate some of Sharepad’s functionality in determining whether a company is cheap in context of the cycle.

Screening for Bargains

On the whole, one ought to be careful about using screens to find good investments. This is especially true if one is using the screens to find bargains since frequently, shares are cheap for a very good reason. Where screens are better is for finding ‘good’ businesses. Consequently, the screen created below ought only to be used to help work out whether something you already like is cheap enough. In other words, it presupposes that you already ‘know’ the businesses. Starting with valuation feels like putting the cart before the horses. The caveat therefore is that the following screen should only be used to either prompt an investor to conduct due diligence on a potential equity investment or, monitor the cyclical cheapness of equities that an investor is already familiar with.

Investing in the Cycle

All investors know that one of the keys to successful investment is not paying too much, or better still, buying something cheap – the value principle. This often takes the form of an earnings-based ratio. However, this inherently assumes that the earnings of a business are ‘correct’, which means that the business in question can sustainably generate commensurate levels of profit. This is not always the case, sometimes a business is either underearning (see Dewhurst now) or overearning (see Halfords c. 2021).

Therefore, looking at an earnings-based metric only provides half the picture, the other half is revealed by assessing the sustainability of the earnings. An insight into this sustainability can be provided by viewing business returns proportional to the assets or capital of a business. Frequently, the use of measures like return on capital employed (ROCE) or cash return on capital invested (CROCI) helps an investor formulate an idea of the quality of a business. That is to say, higher means better quality because it implies the presence of some special sauce like a brand, network, unique asset or monopoly.

The Filter

The purpose of the filter is twofold; to look at the cheapness of a stock on an earnings-based metric compared to its long-term average and to look at what current earnings compare to its long-term average. Long-term here means ten years.

The exact formulations are:

  • EV/EBIT as a percentage of long-term average EV/EBIT – that is to say how cheap is a stock relative to its own history.
  • ROCE, which is EBIT return on Capital Employed as a percentage of long-term average ROCE – that is to say to what extent is it over or under earnings compared to history?

The hope here is that there are intrinsic levels of valuation and profitability toward which an individual stock gravitates.

To create these filters, you right-click on the column header and click ‘Add another column…’

This brings up the following screen, on which you click the big blue button at the bottom left ‘Combine items>>’

That opens up the following box, which allows for the ability to create columns based on formulas.

For our purposes, in the first select box you input ‘ROCE’, in the operator column, set it to ‘% of’.

In the right-hand selection box, once again select ROCE, but this time change the calculated as ‘Average over 10y’.

Repeat the same steps for EV/EBIT. After which the columns will look like this.

We are now ready to perform a screen on cheapness through the cycle. This means we wish to only look at companies that have both a ROCE that is below the long-term average and an EV/EBIT that is below the long-term average. This is achieved by setting the filters on the columns so that both are a maximum of 100%.

This filters the FTSE All Share down to the 73 stocks (as of 23rd December 2023) that are cheaper than the long-term average and under-earning compared to their long-term average. This however will leave several companies that are nearly at long-term averages such as Inchcape above, therefore we can further restrict the list by lowering the maximum levels.

By lowering to 80% in both, the list reduces to just 20 stocks in the FTSE All Share. For completeness, these are listed below and sorted by sector. Please keep in mind that this is just a filter and one should always conduct due diligence on the businesses themselves. Remember – they might be cheap for a reason. As at the end of 2023, fund management firms and stocks related to real estate feature heavily.

Notes

Disclosure: Jamie owns shares in Dewhurst, AB Foods and Renishaw

~

Jamie Ward

Got some thoughts on this week’s commentary from Jamie? Share these in the SharePad “Weekly Market Commentary” chat. Login to SharePad – click on the chat icon in the top right – select or search for “Weekly Market Commentary” chat.