Bottleneck thinking

11.08.2017 By

While getting set-up with a gym membership recently, the salesperson asked a question that made me pause:

“What is the barrier to you signing up today?”

It was a really good question and a good example of bottleneck thinking—namely, that you must first focus on the limiting factor, the part of the system that is most restrictive to the overall system’s output (i.e. the bottleneck).

Bottlenecks can often be spotted by looking for queue formation (queues of people, inventory, information, etc.). For a familiar example, if the security screen at the airport is the slowest part of getting on a flight, there is little benefit in speeding up any of the other steps leading to that moment (e.g., the check-in process), as any extra time created will only result in more waiting at security—i.e., the bottleneck. At the moment of my gym membership sign-up, it was a bit pointless for the salesperson to concentrate on the added benefits of a sauna when one of my primary concerns was how I was going to find parking anywhere nearby.

Even when the system in question is a decision-making process, focusing on alleviating the largest bottleneck is the surest way to speed it up. Our research team has found that the process of identifying bottlenecks can be helpful when discussing potential investments. As an observational lens, bottleneck-thinking can quickly uncover specific pressure points people may have, such as a holding’s valuation or current management.

One of the ways we try to identify so-called bottlenecks is with our quarterly matrix process, whereby we rank all the companies we have invested in across five categories:

  1. Management
  2. Business Model
  3. Risks
  4. Valuation
  5. Skew

The rankings combine into a Quality (business model, management, risks) and Return (valuation, skew) coordinate that we then plot onto a 2D grid (known internally as the “Matrix”). In general, companies that plot high return and high-quality—the upper right area of the matrix—will be weighted more heavily in our portfolios than companies plotted lower left (lower quality and lower return).

Aside from helping us evaluate whether we need to adjust the weights in our portfolio, the matrix process also reveals everyone’s perspective of a company—via a visual tool grounded in a shared language. We can then easily scan each individual matrix score across the five categories to look for the largest deviation.

For example, on our Global Small Cap team, Karan, Christian, Paul, and I may plot the same company in two different locations within our matrix. On closer inspection, it would seem we share the same rankings across all categories with the exception of valuation. We can then fairly quickly narrow-in on what our conversation should address: how the other person thinks about valuation, and the inputs, assumptions, and confidence ranges driving their viewpoint.

By incorporating bottleneck-thinking into such a structured, familiar process to us, we find that our discussions become more focused, and can ultimately move forwards more efficiently.

You Might Also Like

9 Comments

  • Reply
    TG
    11.09.2017 at 11:05 am

    Hello, great article! Just wondering, how do you define skew when defining returns? Thanks

    • mm
      Reply
      John Wilson
      11.20.2017 at 8:17 am

      Please see my response below.

  • Reply
    Brian Clark
    11.12.2017 at 11:00 am

    Good article. The only thing that wasn’t explained well enough (for me) was the term “skew”. What is that a measure of? Would appreciate it if you could advise. Thanks.

    • mm
      Reply
      John Wilson
      11.20.2017 at 8:17 am

      Please see my response below.

      • Reply
        B. Clark
        11.20.2017 at 4:44 pm

        Thanks for the thorough reply. Very helpful.
        BRC

  • mm
    Reply
    John Wilson
    11.20.2017 at 8:16 am

    Thank you for your questions. The value of a company is the sum of all the future cash it spits out to shareholders discounted at an appropriate discount rate. As you can imagine, predicting how much cash flow a company will spit out from next year to eternity is hard to measure precisely. Instead, our philosophy is to create a number of different scenarios (we use a Monte Carlo simulation to accomplish this) that we think the company is likely to encounter. After we do this we step back and look at all the various scenarios and their estimated likelihood. Consistent with our be boring make money approach, we don’t prefer companies where large negative scenarios are likely to occur (think of a company who’s entire profitability rests on one patent or one drug); companies like this would receive a lower skew rating as returns are more skewed downward. Similarly we prefer companies that potentially have very large positive scenarios that could add a lot of value to the shares. We would rank skew higher here. So in sum, skew is our assessment of whether future cash flows of the company are more skewed to the upside or downside.

  • Reply
    TG
    11.21.2017 at 12:23 am

    Thanks again! A follow-up question: the Monte Carlo simulation part seems straightforward, it seems like coming up with a range of scenarios is the most difficult part. How do you decide which scenarios are important and whether your list encompasses all range of possible scenarios?

    • mm
      Reply
      John Wilson
      11.22.2017 at 8:37 am

      You’re absolutely correct in that this is the more complicated part of the process and this is what a large part of our due diligence process attempts to narrow down. This involves a fair amount of research including: understanding cost structures; speaking with competitors and customers to understand their perspective on demand drivers/risks/input costs; and, looking for similar patterns in similar industries or patterns from the same business model in different geographies etc. We also make the task a bit easier on ourselves by focusing on companies that have a barrier to entry or moat that are run by a capable management team, as we believe these companies tend to exhibit more consistent financial performance (they will usually have bargaining power with their customers which allows them to pass changes in costs through to the end customer). And a capable management team can go a long way in protecting a company’s profitability and creating growth avenues for the company.

      As far as your second question, although we believe this research helps to narrow down what we think the likely scenarios are for a particular company, we don’t believe there is any way to know beforehand that we have successfully captured all reasonable scenarios. That’s the nature of forecasting. Instead we try to handle our inevitable errors in this pursuit by 1) reviewing our errors frequently in order to try and determine if there is a pattern in the errors we are making, and 2) diversifying the portfolio and avoiding large concentrated bets.

      • Reply
        TG
        11.22.2017 at 12:58 pm

        Thanks! It is really interesting to know how to tackle these problems in the real world and get a glimpse on the process! Thanks very much!

    Leave a Reply