Be careful of numbers expressed in %
Back in my MBA days I took an elective course in advanced operation management. In one class the professor went through a very detailed case study of an airline company that was suffering from low profit margin. The analysis led us through a series of improvement from 5 or 6 different parts of the entire operation, leading to a 80% (!) increase in the operating profit margin.
80% sounds impressive, right?
Not really. In that case all the proposed efforts increased the nominal operation margin from 0.6% to 1.08% — on paper.
Don’t laugh at the poor profitability to start with just now. Low single-digit operational margin is not uncommon in airline businesses due to the cut-throat competition. There’s nothing wrong about a low margin. The supermarket industry has similarly low margins.
The real problem is that the seemingly impressive 80% improvement is nothing more than a 0.48% net improvement on the actual KPI that matters.
A business consultant will tell you: “But this is huge! If you’re in a low margin business, being able to improve so much will generate a lot more net income for the company given the humongous revenues!”
And he’s surely to claim that, as it’s his business to persuade the CEO of an airline company to pay him and his team fat consulting fees — which is small compared to the annual fixed expenditures of the said airline company but could afford the small team of consultants each a $300k annual salary and to fly business class — to conduct a 6-month project to come up with the said operational improvement plan that would make the CEO look good.
An even worse problem is: none of this kind of plan takes into volatility, let alone the Black Swan events. Even without September 11th, Lehman Shock or Covid-19, the fluctuation of oil prices every year could easily wipe out that 0.48% gain, even if part of the cost is transferred to passengers in ticket prices.
In other words, that 80% is a performance figure expressed in % that is completely meaningless when it comes to measuring the efficacy of this proposed plan.
Reversely, there are some figures in % that look insignificant but can mean tremendous business opportunities, or a path to an unicorn.
A few years ago I was listening to a talk by Taavet Hinrikus, the co-founder of TransferWise with whom we co-invested in a robotics startup together. He was talking about why and how he started TransferWise.
It’s well known that he was the first engineering hire of the Skype team. When he moved to London and Skype was paying his salaries from Estonia, he was troubled by the constantly fluctuating exchange rate and all the opaque hidden fees in the wire transfers. He decided to launch TransferWise to provide transparent, low-cost international remittance services to the digitally savvy.
It might be shocking to some readers but international remittance costs— which include all related fees + the premium charged over spot exchange rates — actually varied dramatically among different combinations of countries and currencies, to the point that the World Bank considered it a humanity problem and has been tracking them and publishing reports every quarter.
When I went through such World Bank reports in 2014 for my own personal research, the total cost of international wire transfer ranged from the more common 3% of the US-EU pair to the 10%+ of Japan-to-EU. Surprisingly the lowest pair back then among G8 countries was actually Russia-to-US, as low as 1% — one wondered why given all the Russian oligarch billionaires tied to Kremlin?
A normal person looking to start such a business would have looked at these prices and thought that if he could offer something cheaper, he would be able to win businesses.
That’s not how Taavet approached the problem. He went further to dig into the internal corporate cost structures of all different wiring methods — bank transfers, Paypal, MoneyGram, Western Union, etc — and strived to build a digital structure whose operational cost is much lower than any of them, so that he could price the services of TransferWise in a way that none of those existing corporates could ever catch up, due to the impossibility for them to completely ditch their internal operational structures and build a new one.
Today on TransferWise the most common pair of international transfer, such as USD-EUR, is priced at 0.40% while the exchange is executed at the spot rate at the moment of transfer. One can verify the instantaneous spot rate by simply Googling it at the time of transfer — I do it all the time. This means that TransferWise has built a fully digital international wire remittance architecture that’s considerably lower than 0.40% of the total amount transferred, true to Taavet’s vision.
I am not an investor in TransferWise so I don’t have any insider info — even if I do this blog would not be where you learn about it — but let’s say that the actual operating cost is 0.20%, giving TransferWise a 50% gross margin. Now, if any engineer finds a new way to reduce that 0.20% to 0.15%, a mere 0.05% difference, say, by figuring how to mass-processing orders in a vastly more efficient way leveraging machine learning, that would mean TransferWise’s gross margin would jump from 50% to 62.5%. That’s a dramatic improvement, coming from a mere 0.05% difference.
Note that unlike the airline example, where a 0.48% improvement in margin would be easily wiped out by external volatility such as oil price fluctuation. Here TransferWise is running on the cloud with enough movability and distributability. Even if a Jihadist decides for whatever reason to crash a Gulf Stream into a data center where TransferWise’s runs a certain part of its business, the system will be back up pretty soon in another data center in no time. This 0.05% improvement is therefore robust against external volatility or even Black Swan events. In this case the humble figure in % translates to a whole lot more in the bottom line, and valuation for that matter.
This is why one should always be careful of numbers expressed in %. It could go either way.
