Why do capital lite outperform

[ratiman]

Everyone knows that capital light lite outperform but why? Let's say there are two business. Same exact growth rate and same FCF. One is capital lite and the other is not. Why does the capital lite typically outperform? The capital lite will have much higher return on capital but it doesn't matter because growth in FCF is the same. My guess is that high capex businesses develop hidden liabilities over time that aren't immediately apparent ie real depreciation is understated.

[Spekulatius]

51 minutes ago, ratiman said:

Everyone knows that capital light lite outperform but why? Let's say there are two business. Same exact growth rate and same FCF. One is capital lite and the other is not. Why does the capital lite typically outperform? The capital lite will have much higher return on capital but it doesn't matter because growth in FCF is the same. My guess is that high capex businesses develop hidden liabilities over time that aren't immediately apparent ie real depreciation is understated.

While it is mathematically correct that if both business have the same growth rate, they should have the same returns, the more Capex heavy business will have to constantly make decisions on Capex and some of those probably won’t work out, Then there could be lumpiness in Capex, which also increases risk as FCF would ebb and flow. All this means that a Capex heavy business in practice has higher variability and higher risk which over time means it is probably not going to perform well. A significant part of the Capex decision I have seen at work had issues with cost overruns, delays or the business changed in a way that the initial assumptions at the start were not correct any more, leading to suboptimal outcomes.

 

A Capex light business does not have these issues so the math being equal means it’s likely a better investment. I think the other issue is that a lot of stocks still traded more on earnings then FCF, so focusing more on FCF (which is harder to determine ) is actually a way to outperform as an investor.

Edited July 27 by Spekulatius

[Eng12345]

@Spekulatius touches on the realities of the scenario. 

 

But I would add that in a make believe scenario where growth and everything else was equal capital light would outperform simply on the basis of cost of capital. In a simplistic view: using money costs money. 

Edited July 27 by Eng12345

[KJP]

1 hour ago, ratiman said:

Same exact growth rate and same FCF. One is capital lite and the other is not. Why does the capital lite typically outperform?

 

I believe your hypothetical is mathematically impossible, and the reason why is why the capital light company will outperform.  Let's assume two companies, Company A earns 50% on equity, and Company B earns 20% on equity, both grow earnings at 20%, both earn $1 in year zero, and both have returns on incremental equity that are identical to their historical ROEs.

 

By hypothesis, Company A's starting equity will be $2 (1/2 = 50% ROE), and to earn $1.20 it would need to reinvest $0.40 of earnings from year zero (.40 * 50% = .20).  It has an extra $0.60 in earnings from year zero that it can use to dividend out, buyback shares, etc.

 

By hypothesis, Company B's starting equity will be $5 (5 * 20% = 1).  To earn and additional $0.20 it needs to reinvest all $1 of its year zero earnings (1 * 20% = .2) In other words, the more capital intensive Company B would have had to reinvest all of its earnings from year zero to obtain the same 20% growth.

 

So, if companies have different returns on equity, they can have the same earnings growth rate, but they will not have the same free cash flow left over after accounting for the earnings that had to be reinvested to achieve that growth.  The difference between the amount of earnings the high ROE company had to invest versus the lower ROE company will be the source of the higher ROE ("capital lite") company's excess returns. 

Edited July 27 by KJP

[LearningMachine]

On 7/27/2024 at 7:52 AM, ratiman said:

Everyone knows that capital light lite outperform but why? Let's say there are two business. Same exact growth rate and same FCF. One is capital lite and the other is not. Why does the capital lite typically outperform? The capital lite will have much higher return on capital but it doesn't matter because growth in FCF is the same. My guess is that high capex businesses develop hidden liabilities over time that aren't immediately apparent ie real depreciation is understated.

 

One thing that is a little harder to understand is that PP&E ends up being a liability not an asset especially with inflation, because when everything doubles in price, that PP&E has to be replaced at double what it cost.  When you start taking into account the life expectancy of PP&E that will need to be replaced, you start realizing a lot of the cash coming in will have to be used to replace PP&E at inflated cost, with much less if anything left for shareholders than you originally thought. 

 

Buffett tried to explain it at http://csinvesting.org/wp-content/uploads/2017/04/Inflation-Swindles-the-Equity-Investor.pdf, but I think he hid the learning between the lines and in riddles instead of making it very clear. 

 

You can work through an example by imagining you have a business that just rents out a $1000 laptop for $300 per year.  On the surface, it sounds great, but if inflation is running at 8.4%, and you have to replace it 5 years later for $1500, all the money you got out of the business goes into purchasing the replacement laptop (assuming you had to keep the cash for future capex instead of investing), with nothing left for you as an owner of the business.  So, a terrible business. 

 

Even if you were to depreciate accurately 20% per year, you would only depreciate $200 per year, making you think that you had $100 in "earnings" each year, but at the end because of inflated cost, there was nothing you got out of the business. 

 

Even worse than you not getting anything out of the business: such a business can easily become a liability instead of an asset if the lifespan of the PP&E ends up being 4 years instead of 5 years that you originally thought.

 

What would you pay for such a business?  Instead of valuing the business as a going concern, I'd be willing to only value it for its liquidation value. 

 

Overall, you can also try to understand this by assuming that PP&E line-item is a liability that is going to grow at the rate of inflation based on the original cost.  For businesses where PP&E is a much a bigger item that Cash flow from Operations, that PP&E replacement liability will grow much faster than Cash flow from Operations because it is a much bigger number than Cash flow from operations for capital heavy businesses.  Over time, that liability will need to be funded out of Cash flow from operations, eating into FCF.

 

The amazing thing to realize is that there are so many businesses that are effectively a form of renting out a $1000 laptop for $300 per year, and Mr. Market continues to value them based on "earnings" instead of liquidation value. 

 

Mr. Market is currently so excited about cloud companies, e.g. Amazon, but when replacement cost starts to show through on PP&E and no cash is able to come out of the business for a long long time, I think Mr. Market will realize eventually. 

 

 

 

 

Edited July 30 by LearningMachine

[Spekulatius]

I think the day people realize that the mighty 6 or seven have become Capex heavy companies, with substantially worse FCF conversion than they used to have,  will be interesting. Right now, everyone seems to believe that they will generate great returns on their AI investments. Even if this were the case, the returns will most likely take quite a few years to materialize and there will be winners and losers.  So far nobody seems to consider the fact that there will be losers. This looks very much like the telecom boom from 1997-2000.

[jfan]

As people have pointed out mathematically and examples from reality regarding difficulties with capex decisions and their respective execution, I am not sure that this simplification regarding capital lite good and capital heavy bad dichotomy is necessarily accurate. I think it depends on the type of product and the level of competition. If competition is absent or minimal (eg railway monopoly or oligopoly), a significant amount of capital invested in real tangible assets that are difficult to replace, all but guarantees a certainty in future cash flows with a very long tail could be quite desirable. Conversely, in many software businesses where it is asset light, but the product is a commodity or the competition is fierce, coupled with the fact that human capital is potentially very expensive, the cash flows may be more elusive than one thinks and the growth runway is shorter than it seems at first glance.

[TwoCitiesCapital]

21 minutes ago, jfan said:

As people have pointed out mathematically and examples from reality regarding difficulties with capex decisions and their respective execution, I am not sure that this simplification regarding capital lite good and capital heavy bad dichotomy is necessarily accurate. I think it depends on the type of product and the level of competition. If competition is absent or minimal (eg railway monopoly or oligopoly), a significant amount of capital invested in real tangible assets that are difficult to replace, all but guarantees a certainty in future cash flows with a very long tail could be quite desirable. Conversely, in many software businesses where it is asset light, but the product is a commodity or the competition is fierce, coupled with the fact that human capital is potentially very expensive, the cash flows may be more elusive than one thinks and the growth runway is shorter than it seems at first glance.

 

+1

This is a great observation 

 

I don't think Buffett is infallible - but there's a reason he's buying utilities, railways, and etc 

Edited July 27 by TwoCitiesCapital

[ratiman]

5 hours ago, KJP said:

 

I believe your hypothetical is mathematically impossible, and the reason why is why the capital light company will outperform.  Let's assume two companies, Company A earns 50% on equity, and Company B earns 20% on equity, both grow earnings at 20%, both earn $1 in year zero, and both have returns on incremental equity that are identical to their historical ROEs.

 

By hypothesis, Company A's starting equity will be $2 (1/2 = 50% ROE), and to earn $1.20 it would need to reinvest $0.40 of earnings from year zero (.40 * 50% = .20).  It has an extra $0.60 in earnings from year zero that it can use to dividend out, buyback shares, etc.

 

By hypothesis, Company B's starting equity will be $5 (5 * 20% = 1).  To earn and additional $0.20 it needs to reinvest all $1 of its year zero earnings (1 * 20% = .2) In other words, the more capital intensive Company B would have had to reinvest all of its earnings from year zero to obtain the same 20% growth.

 

So, if companies have different returns on equity, they can have the same earnings growth rate, but they will not have the same free cash flow left over after accounting for the earnings that had to be reinvested to achieve that growth.  The difference between the amount of earnings the high ROE company had to invest versus the lower ROE company will be the source of the higher ROE ("capital lite") company's excess returns. 

I agree with your analysis but that's why it's FCF and not earnings. The capital heavy business will have higher cash flow and more capex requirements.

[ratiman]

 

Here is a list of highest return stocks over the last 100 years roughly. As you can see there are plenty of capital heavy stocks. The low capital stocks on this list are Altria, S&P, Coke, Pepsi, Hershey, J&J, UST, Tootsie Roll. I didn't include tech or pharma companies because R&D really should be considered capex. So historically capital lite has mainly meant popular brands. There is a reason why CPG companies still trade at 20x multiples. I think what connects all of these companies is that there is no substitute for a tootsie roll when you want a tootsie roll.

 

 

 

 

 

[scorpioncapital]

If you run out of space, you build skyscrapers. If you run out of physical infrastructure, you write books, or get more efficient, or discover an asteroid and go get some resources - that's all 'asset lite'. Intellectual property is quite highly valued short of acute shortages like war taking oil or chips offline as an example. Even that would recover in time.

 

 

[dwy000]

In order to have the same FCF the asset heavy business has to have much higher gross margins than the asset lite business.  That makes it much more susceptible to competition and it usually comes down to which competitor has the lowest cost of capital (or is willing to accept a lower ROI).

[Cigarbutt]

On 7/27/2024 at 10:52 AM, ratiman said:

Everyone knows that capital light lite outperform but...

What if

-"everyone knows that capital light lite outperform"

-"everyone" (the market) determines the multiple to pay for better things to come

Then

-Isn't this initially priced in by "everyone"?

[Intelligent_Investor]

4 hours ago, Cigarbutt said:

What if

-"everyone knows that capital light lite outperform"

-"everyone" (the market) determines the multiple to pay for better things to come

Then

-Isn't this initially priced in by "everyone"?

Its actually very hard to overprice capital light businesses due to the mathematics of exponential growth over time (capital light businesses are generally considered to be compounders so you are supposed to hold for a long period rather than sell in a year or 2). The basic equations of finance and what is considered "expensive" vs "cheap" are based on ratios rooted in linear math, so it breaks down when you compare it vs exponential compounding over a longer timeframe. 40x FCF is cheap for a capital light business if it can compound for several decades

[Spekulatius]

9 hours ago, Intelligent_Investor said:

Its actually very hard to overprice capital light businesses due to the mathematics of exponential growth over time (capital light businesses are generally considered to be compounders so you are supposed to hold for a long period rather than sell in a year or 2). The basic equations of finance and what is considered "expensive" vs "cheap" are based on ratios rooted in linear math, so it breaks down when you compare it vs exponential compounding over a longer timeframe. 40x FCF is cheap for a capital light business if it can compound for several decades

That’s true for a business that requires Capex too, that’s the feature of exponential growth not capital light or Capex heavy. if a Capex heavy business can keep the incremental return on invested capital high, it should do just as well.

[Eldad]

Capital heavy can’t grow without ever increasing PP&E, capital light can. That is why your scenario is impossible. In order to grow it would have to spend on capex lowering free flow.
 

Also, Capital light can turn off fixed costs in a recession, capital heavy can’t.
 

See Charlie Munger talking about his friend that had to keep putting all of his profits into new cranes or dump trucks or whatever it was. 
 

Edited July 31 by Eldad

[Eldad]

CapHeavy raises $200 in capital. It makes a 25% return on capital. Year 1 it makes $50. It thinks it can grow earnings in year 2 to $75. In order to do this it must spend the entire $50 in retained earnings on new PP&E and borrow another $50 ($300 in total capital to get to $75)

Also it’s equipment it bought in year 1 will start to require maintenance capex soon. FCF = 0

 

CapLight raises $100 and makes 50%. It makes $50 the first year and thinks it can make $75 in year 2. It must invest all $50 retained earnings to get to $75 but it doesn’t have to borrow money like cap heavy. FCF = 0

 

Year 3 $90 in earnings for both. 

Cap Heavy needs to reinvest $65 FCF= $10

 

Cap Light needs to reinvest $30

FCF = $45

 

That basically goes on into infinity with Cap Light destroying Cap Heavy. Get out a spreadsheet, it is painfully obvious. 

 

Edited July 31 by Eldad

[Eldad]

10 minutes ago, Eldad said:

CapHeavy raises $200 in capital. It makes a 25% return on capital. Year 1 it makes $50. It thinks it can grow earnings in year 2 to $75. In order to do this it must spend the entire $50 in retained earnings on new PP&E and borrow another $50 ($300 in total capital to get to $75)

Also it’s equipment it bought in year 1 will start to require maintenance capex soon. FCF = 0

 

CapLight raises $100 and makes 50%. It makes $50 the first year and thinks it can make $75 in year 2. It must invest all $50 retained earnings to get to $75 but it doesn’t have to borrow money like cap heavy. FCF = 0

 

Year 3 $90 in earnings for both. 

Cap Heavy needs to reinvest $65 FCF= $10

 

Cap Light needs to reinvest $30

FCF = $45

 

That basically goes on into infinity with Cap Light destroying Cap Heavy. Get out a spreadsheet, it is painfully obvious. 

 

Should be $60 in capex and $15 in FCF for CapHeavy after year 2. 

[Eldad]

On 7/27/2024 at 11:16 AM, KJP said:

 

I believe your hypothetical is mathematically impossible, and the reason why is why the capital light company will outperform.  Let's assume two companies, Company A earns 50% on equity, and Company B earns 20% on equity, both grow earnings at 20%, both earn $1 in year zero, and both have returns on incremental equity that are identical to their historical ROEs.

 

By hypothesis, Company A's starting equity will be $2 (1/2 = 50% ROE), and to earn $1.20 it would need to reinvest $0.40 of earnings from year zero (.40 * 50% = .20).  It has an extra $0.60 in earnings from year zero that it can use to dividend out, buyback shares, etc.

 

By hypothesis, Company B's starting equity will be $5 (5 * 20% = 1).  To earn and additional $0.20 it needs to reinvest all $1 of its year zero earnings (1 * 20% = .2) In other words, the more capital intensive Company B would have had to reinvest all of its earnings from year zero to obtain the same 20% growth.

 

So, if companies have different returns on equity, they can have the same earnings growth rate, but they will not have the same free cash flow left over after accounting for the earnings that had to be reinvested to achieve that growth.  The difference between the amount of earnings the high ROE company had to invest versus the lower ROE company will be the source of the higher ROE ("capital lite") company's excess returns. 

This is correct. Sorry didn’t see. 

[Spekulatius]

I think @KJP explanation is correct generally. However, there are cases where a business requiring Capex to scale have very high return on assets. examples are some widget makers like SSD or some medical equipment companies.. Very few will grow 20% though. Wr would not define them as Cpex heavy either, but they do require Capex to grow.

 

On the other hand, we have a lotmof Capex light business that have all that high Returns on Capital. The most common reasons are that thy keep cash on the balance sheet or do acquisition and acquire intangibles.

 

@ratiman mentioned same growth rate and if it is organic, then I agree the capital light business should see thru ROE go to infinity theoretically over time. In practise,  that's not the case and growth will almost always be a mix of organic and inorganic growth where the latter puts intangible on the balance sheet that reduces ROE.

 

I think we are seeing an interesting case of a shift with tech companies playing out where traditional capital light software cos become much more Capex heavy due to a mix change towards cloud and AI which require expensive data centers to run.

[ratiman]

It's a good point that a capital lite business can grow to the sky but a capital heavy business is constrained by real resources. If a cap heavy business compounded indefinitely it would eventually consume all of the real resources but that constraint does not apply to software. If the high capex business did succeed in producing real resources at near zero it would put itself out of business, like nat gas producers.

[KJP]

15 minutes ago, Spekulatius said:

I think @KJP explanation is correct generally. However, there are cases where a business requiring Capex to scale have very high return on assets. examples are some widget makers like SSD or some medical equipment companies.. Very few will grow 20% though. Wr would not define them as Cpex heavy either, but they do require Capex to grow.

 

 

The opening post was a bit blurry because it asked about what tends to happen in the real world and about a hypothetical that appears to contain a false premise.

 

My explanation is only a response to the hypothetical posed in the opening post:  Why will a "capital lite" company outperform [I assume this means produce higher total shareholder returns] a capital intensive company if they have the same growth rate and FCF?  If you add an assumption that each company's incremental capital earns each company's historical returns, then the hypo is impossible because true FCF after capital reinvested for growth cannot be the same for both companies.  That must be true given the hypothetical's assumptions about RoEs and growth rates.

 

Most of the responses in the thread seem to get at the other question raised in the opening post:  In the real world, why do capital lite companies tend to outperform over time?  I think you're right that that statement is not universally true and that the key questions for the long-term return on any growing business are:  "How much can the company reinvest (or acquire) and at what rates of return?"

[Eldad]

1 hour ago, KJP said:

 

The opening post was a bit blurry because it asked about what tends to happen in the real world and about a hypothetical that appears to contain a false premise.

 

My explanation is only a response to the hypothetical posed in the opening post:  Why will a "capital lite" company outperform [I assume this means produce higher total shareholder returns] a capital intensive company if they have the same growth rate and FCF?  If you add an assumption that each company's incremental capital earns each company's historical returns, then the hypo is impossible because true FCF after capital reinvested for growth cannot be the same for both companies.  That must be true given the hypothetical's assumptions about RoEs and growth rates.

 

Most of the responses in the thread seem to get at the other question raised in the opening post:  In the real world, why do capital lite companies tend to outperform over time?  I think you're right that that statement is not universally true and that the key questions for the long-term return on any growing business are:  "How much can the company reinvest (or acquire) and at what rates of return?"

Capital Heavy can still outperform if it has a higher growth rate. All that matters is growth and return. 
 

But Cap heavy has so many disadvantages they are likely to give it all back over time to a slow and steady cap light IMO.