I want to talk about an interesting dilemma that afflicts investors – or at least, afflicts me. In my earlier post ‘Is Thorn Group Ethical?‘ I theorized at some length about the company’s leasing practices and their customers.
Let me say here that it is important to understand how a business works, before you buy shares in it.
But consider this paradox. This is one thing about investing that drives me batty:
You can spend a lot time doing research that does not improve your predictive chance of success in a stock, but you could lose badly if you do not do it.
Now maybe I’m over-complicating this but in my mind there is always a tradeoff between the amount and type of research that you do, and the actual value added to your thesis by doing that research. This is a fairly weighty post, but here are my general thoughts on the issue.
(If you die a little when confronted with a wall of text, just skip ahead to the TL:DR section at the end).
Imagine that you wanted to make sure that Thorn Group was leasing to customers who were using its services to buy essentials (and thus that demand was defensive), not risky credits buying Playstations. You could spend a long time figuring out how many items from the ‘essential’ (fridge, washing machine) and ‘nonessential’ (Playstation, stereo system) categories that Thorn sold. You could spend even longer figuring out if it was Centrelink recipients or other customer categories that bought them.
And at the end of the day, all you would know is the # of demographic X that bought essential or nonessential items. You wouldn’t really understand the drivers of those purchases (although you could theorize, maybe accurately). If you tracked changes in the data over time, you would have a good shot at identifying problematic correlations or changes in relationships between variables. But you still wouldn’t know if those relationships were going to change. I.e., you would know what you had, not what was going to happen.
Imagine that I discover that 40% of Thorn’s sales come from irresponsible spenders with 3 maxed out credit cards and no access to other finance. This is obviously not ideal for the shareholder. However, given what Thorn does and the environment in Australia over the past decade, it’s likely that this type of customer has (in this theoretical example) been a fixture at Thorn for the past decade. Should I sell?
On many occasions since the GFC, Australian banks have looked much the same as this theoretical Thorn – lending a lot of money to people with..well, with ‘obvious’ red flags against them (ahem, million dollar shack in rural mining town, or on a macro level, record household debt to GDP ratio). But the fact is that selling the banks at most points during the past decade has been a poor investing decision. Not because of the risks, but because the drivers of the earnings growth have not changed (low cost of debt, stable employment, rising house prices, etc).
So potentially as an investor you could spend a long time understanding what’s going on in a business, without actually improving your predictive ability for determining whether you’re going to make money.
To give an example:
If I make money in Thorn, it will be because the class action is a fizzle, there are no further regulatory penalties, customers continue to use the company’s services at a similar rate (with similar levels of profitability and bad debts etc), and the share price normalises on a higher P/E multiple.
For simplicity’s sake, I will focus on just one issue for my example here: Customers continuing to use Thorn’s services at a similar rate.
Imagine that every last one of Thorn’s customers is financially hopeless and has 3 maxed out credit cards. The real question is not ‘are these customers potential bad debts‘ – of course they are. In my earlier article I established my belief that anybody with alternative sources of finance does not use Thorn’s services.
The real question is ‘what are the drivers of demand for Thorn’s services?‘ And in that sense, the demographics of the customer are not as relevant as other questions like ‘what are the alternative sources of finance for these purchases, that could see customers not buy from Thorn?‘
It’s partly a matter of employing second-level thinking, but also partly focusing on the most fruitful avenues for your inquiries. Because time and mental energy are finite. This leads me neatly to part 2 of my sermon: Heuristics.
But first, an intermission. Go and watch some videos on youtube. Here’s one of my favourite easy listening tracks right now, by the most promising up + comer in country music:
I shall proceed.
Part 2: Heuristics
So, you’ve spent lots of time diving down a rabbit hole to understand what’s going on in a business, and you encounter several potential outcomes:
- You find problems with the business
- You don’t find problems with the business
- You find something, but you either a) think it is a problem when it is not, or b) think it is not a problem when it is
- You find shades of grey, where you find a problem (or a point of strength) and think a) it is worse (/better) than it actually is or b) not as bad (/good) as it actually is.
Sometimes you can very clearly understand what is going on (or at least you think you do), because of your experience or for other reasons. But show me an investor that sails through the above dilemma 100% every time on every company and I will show you a liar. The potential for stumbles is high.
So you can spend all this time researching a factor that you think is a concern (customer demographics and financial position) only to find out much later that it is not as significant as you thought, or that you got tripped up by some other factor like debt, competition, regulation, or unforeseen risks. You have to multiply the investigative process many times over to check out all relevant factors, which is time consuming and exhausting.
So a majority of investors use heuristics (‘mental shortcuts’) to cut corners. There are so many unknowns in investing, that there is no reasonable way for active investors to avoid the use of heuristics.
Given the choice between interviewing 1000 Thorn customers and getting honest answers to whatever questions I asked, and being able to view a detailed breakdown of Thorn’s sales by category and to which demographic (eg customer avg income, financial position), I believe the former would give me a better understanding of the drivers of the business’ earnings. This should add more value to my investing thesis because it is earnings that drive investment results.
Yet, if you gave me the choice, I would actually (perversely) choose the latter, because it allows me access to heuristic information that would save me a lot of time, cost, and effort. Yet as I said earlier, knowing this information does not explain the why behind the demand for Thorn’s services. Information that I draw from the demographic data would be a heuristic, eg: “customers with 3 or more credit cards are irresponsible spenders, there are a lot of customers with 3+ credit cards, and thus demand for Thorn’s products is not defensive in event of an economic shock“.
If you tried to draw that conclusion from that data in a hard science, you would be dragged out back and shot. But at the same time I can tell you accurately from a common sense perspective – and I suspect most readers would agree – that most people on an ordinary income with 3 maxed out credit cards are not financially responsible. So investors can safely and valuably use some heuristics at least some of the time.
There is an additional lens to see this issue through. You need to make sure that your heuristics line up with the actual risk you’re trying to evaluate. For example, there is an overlap between the amount of debt an individual has (3 credit cards) and their ability to repay, but it’s not a perfect 1:1 correlation, because it overlooks the potential of all sorts of other unforeseen factors or disruptions (e.g. unemployment, divorce, changes in the cost of finance, to name the obvious ones). So I could spend all this effort investigating Thorn’s customers financial position when really what I should have been looking at is factors that affect their ability to repay and/or the likelihood of economic/regulatory/other shocks occurring that could affect it. I’d almost be better off reading Reserve Bank employment numbers and industry publications on trends in consumer leasing than I would actually looking at Thorn customers.
And so after we swing through that big old circle, we’re right back at the start where you could very effectively use a heuristic and simply track Thorn’s credit impairment % over time, as this implicitly covers a majority of the above risks, and pick an upper limit or a trend at which point you would freak out and sell. This figure is available in the annual report and in a sense, why would you even bother looking further?
But. But but but. You can’t do that because companies can do some pretty freaky things to keep business looking good. So every investor has to weigh up how much investigation they are going to do against the potential to improve the predictive value of their thesis (either in terms of likely upside, or downside risk).
One final very quick example (I promise)
Another place this happens is with ‘tailwind investing’. You can do a lot of research into a tailwind to determine, just say, that an industry is growing at 10% per annum forever. But that’s a big waste of time because even if you’re totally wrong and the industry grows at -1%, there will still be winning companies in that industry, winning due to other factors (such as superior competitive position or consolidation) that are more important/predictive.
Too Long: Didn’t Read (TL:DR)
In summary, it is very easy in investing to get sucked into a big old black hole that a) wastes a lot of your time, b) misses the point, and c) doesn’t improve your chances of success in a stock. You will save a lot of time by sticking to the basics.
And if you stick to the basics (e.g. what’s in the reports), I can almost guarantee that you’ll get suckered sooner or later by some great situation you would have avoided if you’d spent 10 minutes reading a government, industry, or even a media publication about a company’s main line of business or its customers.
This is something that I grapple with a lot. I feel like I am perpetually balancing:
- the benefits of heuristics against the need for quality information, bearing in mind that time and energy are finite, and
- the value added by focusing on the most important elements of the thesis, against the risks of missing something important in a seemingly less relevant area
- the investment of time + effort against the likelihood of actually improving my chances of investment success (or reducing risk of failure)
And I don’t really have a solution, other than to err on the side of caution. I would be interested to hear reader thoughts on the matter.
Note: I use Thorn purely as an example. Yes this dilemma applies to Thorn, but it also applies to every investment. I have used Thorn simply as a vehicle to help me frame my thoughts, and there are no changes to my Thorn thesis.
I own shares in Thorn Group.