Woolworths - How to decide when to leave a party?
This is a longer version of an article originally published in the Australian Financial Review.
Within investing, dealing with success isn’t the problem; understanding when your road-map isn’t working out is the hard part. If you change your mind too early, you run the risk of shutting down good ideas on what we’d call “noise”. This may be a supplier whose view is given too much weight; it may be the share price that moves the wrong way causing doubts; or it may be a news story that isn’t that relevant.
On the other hand, waiting too long can see either all your profits lost, or conversely large losses emerge before you admit an error. You are looking to separate the “signal” from the “noise” to use Nate Silvers book title.
One way of approaching this problem is to use a model called “ Analysis of Competing Hypotheses”. It’s a model that the Central Intelligence Agency use to help field officers on this signal and noise problem. It’s based around the idea that pieces of data are compatible with multiple theses, but if you have a single thesis you find data will fit that thesis, convincing yourself that you are on the right track, when you have never considered other theses. It’s a handy way to look at problems with incomplete information.
The case of Woolworths
This brings us to Woolworths, which reports earnings next week. The original thesis in 2015 was of a stock that was over-earning with margins that were too high. This thesis has played out over the intervening period to such a point where we’d acknowledge that the idea has become a “consensus short”.
The idea was that there would be a second leg to the story with a competitive response from Wesfarmers taking pricing down to protect Coles’ newly grown market share gains. The hypothesis being that Coles would not let Woolworths turn around easily, learning from the mistakes that Woolworths made in the 2009-13 period. This would lead to lower margins across the industry with the growth of Aldi into Western Australia and South Australia exacerbating problems.
What emerged before Christmas was that Woolworths was doing better on sales. This was acknowledged as expected given the price cuts, but the view was that pressures would show up in margin pressures. In early 2017 more news came in of suppliers suggesting that Woolworths did better in the Christmas trading period. But this time there were also reports that suppliers have found them better to deal with and are seeing better volumes. A cross in our hypothesis column.
The news that switched the balance of probabilities was that Coles has decided to not push back on pricing. In essence it is going to cede margin and share to Woolworths and hope to get margin back through cost out initiatives. This put another cross against the thesis, but, even more, it suggests that an alternate hypothesis – that is that Woolworths is fine for now – has more “ticks” than the original hypothesis. It also means that margins have probably bottomed in this half at a number higher than 4% in the Food and Liquor business.
Then is the stock a buy?
A very good question. Because logically it seems to make sense: if the stock isn’t going down, it must go up? Here we applied another tool that helps analyse these questions, but this time it’s more numerically based. We’d call it “what do you need to believe for X to be true?” but the more common name in the industry is a “reverse DCF”. Discounted Cashflow (DCF) analysis takes your assumptions on earnings to derive a valuation based on those future earnings. A reverse DCF on the other hand, takes the market price as the correct price and then re-engineers the assumptions to make the price be constant.
As a result you get a series different assumptions in tables that hold true at certain prices. Why is this interesting? Because you can see whether you find those assumptions to be plausible and test them against what you think about the world.
So let’s walk through the Woolworths assumptions. The first black boxed areas are the current market assumptions for FY-18 at the current share price ($1.20 EPS and 20x P/E). The red boxes are the payoff structures where a short can make more than 10-15%; the green is for longs to make more than 15% from here. As you can see there are a lot of permutations around what multiple the market choses to pay for the stock, as well as the EPS.
The blue values I have chosen are one payoff structure that is more plausible in our minds: to make the stock worth $29, and keep the P/E constant, you need to believe the business can earn $1.40 of EPS in FY-18.
Figure 1 - Woolworths payoff scenarios
The next table then takes the $1.40 of EPS and shows the combinations of Food and Liquor EBIT margin and Food and Liquor sales growth to get to that $1.40 (note we are using Morgan Stanley forecasts as the base case and assuming the other businesses outside of Food earn MS’ FY-18 estimates).
As per before there are differing payoffs, but one that we look at here (blue again) is that the business earns EBIT margins of 5.4% and sales grow at 4.5%. Note how there are few payoff combinations when using FY17 margins and sales as the mid-point.
Figure 2 - Combinations of Food and Liquor EBIT margin and Food and Liquor sales growth to get to $1.40 of EPS
The last step is to test the sales assumptions here, by breaking it down into price and volume. So we take the 4.5% and look at the combinations. Given population growth in Australia is circa 1-1.5%, you would expect industry volume to grow in that range, and so it implies pricing inflation of 2.6% in FY-18.
Figure 3 - Combinations of change in volume and in price to get to sales growth of 4.5%
What does this all mean?
What became abundantly clear when we were doing this is that the stock is neither a buy nor a sell currently if the view that Coles doesn’t push back on pricing holds. This may disappoint people, but it’s the reality of most stocks on the market – they are relatively fairly priced. What also became clear is that at prices north of $28, the stock is baking in some heroic assumptions on margin turnaround and pricing in the industry.
To get really bearish on the stock, given what we have learnt in the last three months, the stock undergoes a sudden de-rating of its P/E which seems implausible given it didn’t de-rate when earnings were missing badly. There are too many investors who “want to believe”.
What were your actions?
So the short was covered two weeks ago ahead of the upcoming results. Our view is that Coles is taking the wrong lesson from the UK experience – that they have the wrong “mental model”. Wesfarmers believes the UK price war was irrational, when it was a rational price war, as in fixed cost leverage businesses it’s rational to try and hold share (see Qantas and their “line in the sand”).
When this Team Morphic member was younger, there was a saying “nothing good happens after 3am”. We think 3am may have passed for now in the stock. Time to leave the party.