Thread by @10kdiver

1/

Here's the thing about Return On Equity (ROE).

Imagine we have a business with $1B in assets, that earns $200M each year on these assets.

The business is stable and has durable competitive advantages. So, it can sustain these earnings year after year.

2/

Let's say the business has $800M in debt (on which it pays 5% interest) -- and no other liabilities.

So, after paying 5% of $800M = $40M in interest, the business will post $160M in earnings on $200M in equity, for an 80% ROE.

3/

Suppose the business now uses its earnings to pay down part of its debt.

This leaves it with the same $1B in assets, but only $800M - $160M = $640M in debt.

Equity therefore goes up from $200M to $1B - $640M = $360M.

4/

What happens to ROE in Year 2?

It turns out: ROE will drop precipitously -- from 80% to just ~46.67%.

That's despite the fundamental earning power of the business not having changed at all! (The business still earns the same $200M pre-interest on $1B of assets.)

5/

And if the business keeps this up (earning the same $200M pre-interest, and paying down debt each year), ROE will keep dropping year after year.

6/

To those who use ROE as an indicator of business quality, it may *look* like the business is steadily becoming worse.

But reality is very different. The business's earning power is in tact. And as debt is paid down, the business arguably becomes *less* risky, *less* fragile.

7/

It's just that the capital structure of the business is shifting.

The equity/debt mix is tilting more towards equity than debt.

But that causes ROE to drop.

8/

If management wants to increase ROE, they can do the hard things: improve operating margins, increase returns on invested capital, etc.

Or they can do the easy thing and simply saddle the business with ever-increasing levels of debt.

9/

That's why Buffett likes to focus on the return a business earns on *unleveraged* net tangible assets.

This is much harder to "game" by simply playing around with the debt/equity capital mix.

10/

So, the bottom line is: when studying changes in a company's ROE over time, it's important to understand *how* those changes were brought about.

Were they the result of genuine operating improvements, or simply the after effects of issuing imprudent amounts of debt?

/End