Read on Twitter
1/
Get a cup of coffee.
In this thread, I'll show you a simple way to understand and analyze stock buybacks.
Get a cup of coffee.
In this thread, I'll show you a simple way to understand and analyze stock buybacks.
2/
Imagine that you're a part owner of a pizza shop.
The shop has 20 owners. You and 19 others. Each of you owns 5% of the shop.
Imagine that you're a part owner of a pizza shop.
The shop has 20 owners. You and 19 others. Each of you owns 5% of the shop.
3/
Suppose the shop made $1M in 2019.
A simple way to distribute this profit would be to give each owner $50K:
(20 owners) * (50K/owner) = $1M total profit.
This is a "dividend policy". Each owner gets a dividend whether they want it or not.
Suppose the shop made $1M in 2019.
A simple way to distribute this profit would be to give each owner $50K:
(20 owners) * (50K/owner) = $1M total profit.
This is a "dividend policy". Each owner gets a dividend whether they want it or not.
4/
There's also another way to distribute the profit.
Some owners can use their share of the profit to buy out other owners.
This is a "buyback policy". Owners who don't need cash can forego their share of *current* profits in exchange for a greater % of *future* profits.
There's also another way to distribute the profit.
Some owners can use their share of the profit to buy out other owners.
This is a "buyback policy". Owners who don't need cash can forego their share of *current* profits in exchange for a greater % of *future* profits.
5/
Let's say pizza shops in your area are worth 10 times earnings.
So this pizza shop is worth about $10M at the end of 2019. Each owner's stake is about $500K.
So the $1M 2019 profit can be used to buy out 2 of the owners, assuming the other 18 are OK with that.
Let's say pizza shops in your area are worth 10 times earnings.
So this pizza shop is worth about $10M at the end of 2019. Each owner's stake is about $500K.
So the $1M 2019 profit can be used to buy out 2 of the owners, assuming the other 18 are OK with that.
6/
Let's say it works out. You and 17 other owners get together, and buy out the other 2 owners using the $1M.
Now, the shop has only 18 owners, not 20.
And each owner now owns roughly 5.55% of the shop, not 5%.
Let's say it works out. You and 17 other owners get together, and buy out the other 2 owners using the $1M.
Now, the shop has only 18 owners, not 20.
And each owner now owns roughly 5.55% of the shop, not 5%.
7/
Suppose the shop sells more pizza in 2020, and therefore makes $1.1M: a 10% increase over 2019.
Now, the shop as a whole is worth $11M. Same 10% increase. At the same earnings multiple of 10.
But each owner's stake is now worth $611K ($11M/18), a 22% increase!
Suppose the shop sells more pizza in 2020, and therefore makes $1.1M: a 10% increase over 2019.
Now, the shop as a whole is worth $11M. Same 10% increase. At the same earnings multiple of 10.
But each owner's stake is now worth $611K ($11M/18), a 22% increase!
8/
Key lesson (the magic of stock buybacks): Over time, stock buybacks can make your investment compound at a rate *higher* than what the underlying business is growing at.
Also, buybacks are more tax-efficient than dividends.
Key lesson (the magic of stock buybacks): Over time, stock buybacks can make your investment compound at a rate *higher* than what the underlying business is growing at.
Also, buybacks are more tax-efficient than dividends.
9/
Public companies are a bit more complicated than this pizza shop example.
With public companies, the main variables that affect buybacks are: Free Cash Flow, Growth, Capital allocation, Dilution, and Multiple expansion/contraction.
Public companies are a bit more complicated than this pizza shop example.
With public companies, the main variables that affect buybacks are: Free Cash Flow, Growth, Capital allocation, Dilution, and Multiple expansion/contraction.
10/
Free Cash Flow (FCF). Only companies that generate lots of predictable FCF year after year can sustain meaningful share buybacks.
Growth. Companies that grow FCF year after year can do larger buybacks over time.
Free Cash Flow (FCF). Only companies that generate lots of predictable FCF year after year can sustain meaningful share buybacks.
Growth. Companies that grow FCF year after year can do larger buybacks over time.
11/
Capital allocation. Companies have many ways to deploy FCF. They can do buybacks. Or pay dividends. Or make acquisitions. Or simply pile up cash on the balance sheet.
It is up to management to intelligently decide how much FCF to allot to each of these activities.
Capital allocation. Companies have many ways to deploy FCF. They can do buybacks. Or pay dividends. Or make acquisitions. Or simply pile up cash on the balance sheet.
It is up to management to intelligently decide how much FCF to allot to each of these activities.
12/
Dilution. The purpose of buybacks is to increase shareholder returns by reducing the number of shares outstanding over time.
But many companies negate a lot of this benefit by simultaneously issuing vast amounts of stock to executives, via share based compensation (SBC).
Dilution. The purpose of buybacks is to increase shareholder returns by reducing the number of shares outstanding over time.
But many companies negate a lot of this benefit by simultaneously issuing vast amounts of stock to executives, via share based compensation (SBC).
13/
Multiple expansion/contraction. Share buybacks are more effective at cheaper prices (more shares retired for the same dollar amount spent).
But many companies are poor timers of their own stock. They do buybacks with gusto at high multiples and halt them at low multiples.
Multiple expansion/contraction. Share buybacks are more effective at cheaper prices (more shares retired for the same dollar amount spent).
But many companies are poor timers of their own stock. They do buybacks with gusto at high multiples and halt them at low multiples.
14/
Let's put some numbers to these variables.
Say XYZ Corp. has $1B of FCF that grows at 10% per year. They spend 30% of FCF on dividends and 70% on buybacks. Shareholders are diluted 1% per year via SBC. The stock trades at 20 times FCF.
Let's put some numbers to these variables.
Say XYZ Corp. has $1B of FCF that grows at 10% per year. They spend 30% of FCF on dividends and 70% on buybacks. Shareholders are diluted 1% per year via SBC. The stock trades at 20 times FCF.
15/
What return can an investor expect from the stock over the next 15 years?
There's a simple way to solve such problems: assume a certain share count (say 1B) at the outset. Then work out what happens to the share count, FCF, the stock price, etc., over time.
Like so:
What return can an investor expect from the stock over the next 15 years?
There's a simple way to solve such problems: assume a certain share count (say 1B) at the outset. Then work out what happens to the share count, FCF, the stock price, etc., over time.
Like so:
16/
In the example above, the buybacks transformed our 10% growth company into a 14.51% compounder over 15 years.
But P/FCF multiples rarely stay constant for this long. What if the multiple slowly increases from 20 to 40 over this period?
Now the return increases to 18.22%:
In the example above, the buybacks transformed our 10% growth company into a 14.51% compounder over 15 years.
But P/FCF multiples rarely stay constant for this long. What if the multiple slowly increases from 20 to 40 over this period?
Now the return increases to 18.22%:
17/
What if the multiple decreases from 20 to 10 over this period?
Now the return drops to 11.4%:
What if the multiple decreases from 20 to 10 over this period?
Now the return drops to 11.4%:
18/
That's how you build a buyback model that accounts for growth, capital allocation, dilution, and multiple expansion/contraction.
You should then run the model under various assumptions. Make sure the upside is sizable and the downside is tolerable -- *before* investing.
That's how you build a buyback model that accounts for growth, capital allocation, dilution, and multiple expansion/contraction.
You should then run the model under various assumptions. Make sure the upside is sizable and the downside is tolerable -- *before* investing.
19/
There's also a back-of-the-envelope quick and dirty approximation you can use to gauge a buyback without going through the model above.
But this is only a rough approximation. So please use with caution:
There's also a back-of-the-envelope quick and dirty approximation you can use to gauge a buyback without going through the model above.
But this is only a rough approximation. So please use with caution:
20/
The formula above says: if you hold the stock for a long time (i.e., n is large), then CAGR is inversely related to the average multiple at which buybacks are done.
This means, after buying the stock, you should hope for the price to stay low, not rise!
The formula above says: if you hold the stock for a long time (i.e., n is large), then CAGR is inversely related to the average multiple at which buybacks are done.
This means, after buying the stock, you should hope for the price to stay low, not rise!
21/
Buffett says exactly this in his 2011 letter: https://www.berkshirehathaway.com/letters/2011ltr.pdf
Buffett says exactly this in his 2011 letter: https://www.berkshirehathaway.com/letters/2011ltr.pdf
22/
This @FocusedCompound episode also has a good discussion of the capital allocation behind buybacks.
Andrew and Geoff talk about situations where cash on hand can be more valuable than using it for buybacks, PR problems that buybacks can create, etc.
This @FocusedCompound episode also has a good discussion of the capital allocation behind buybacks.
Andrew and Geoff talk about situations where cash on hand can be more valuable than using it for buybacks, PR problems that buybacks can create, etc.
23/
Share buybacks can be confusing. They're poorly understood and constantly vilified.
I hope this thread was able to shed some light.
Thanks for reading. Enjoy your weekend!
/End
Share buybacks can be confusing. They're poorly understood and constantly vilified.
I hope this thread was able to shed some light.
Thanks for reading. Enjoy your weekend!
/End
