Evaluating REITs
Recently we’ve talked about McDonalds, a fast-food franchiser with a business model that heavily involves owning properties, and then leasing them back to their franchisees.
At the time, I mentioned how this particular business model gave it an incredible stability in revenues compared to a simple restaurant operator business model, but I didn’t really go in depth into that.
Part of that is because I personally am not entirely sure how to evaluate REITs, and the pitfalls involved with them.
This is a big problem for me, since I’m heavily invested in a REIT, Realty Income!
The problem I have with REITS
REITs are special.
To be more precise, there are many people that say that due to the business model and tax differences in the business the standard valuation metrics do not apply to REITs.
In particular the claim that “Net Income is Not a Good Valuation Metric” is a common statement.
Many people say that the high non-cash charges involved with depreciation and amortization should be removed from the calculation, since Real Estate tends to appreciate.
This is a very popular statement, and indeed many investors an analysts recommend evaluating REITs by their AFFO.
If you’re not familiar with AFFO, that’s because its a non-standard non-GAAP measure. Though there is no one official formula, calculations for AFFO typically would be something like:
AFFO = FFO + rent increases - capital expenditures - routine maintenance amounts
Do you see the problem here?
Let me list out the issues I have here:
Non-GAAP measures are always a bad thing since they are inconsistent and non-standardized.
The measures used to calculate AFFO, namely Capital Expenditures and “Routine Maintenance” have so many holes and so much leeway that management can push anything through this Swiss cheese.
Depreciation and Amortization are real expenses, and discarding them in favor of a vague and easily manipulated proxy is a very bad idea.
That last part is really a sticking point for me.
Depreciation and Amortization are real expenses that come out of your bottom line, and you ignore them at your own peril.
I’ve gone over this before when I talked about the One Man Corporation:
Depreciation is the loss in value that your assets suffer over time and use. Amortization is similar to Depreciation.
In this case we need to depreciate our Axe as well as our right to chop trees.
The reason for this is simple, and is most easily seen with the right to chop trees. At the beginning of the month we had 12 months of tree chopping rights, but now we only have 11. This means we lost 1/12th of its value, or $83.33, because we’ve used up 1/12th of its lifespan.
The case for the axe is similar, even though the Axe doesn’t have a set expiration date like the Right to Chop Trees, that axe will lose value over time. It will chip and become less effective over time, and eventually it will break.
The same things that applied to this Tree-Chopping corporation, also apply to REITs, regardless of whether investors choose to ignore it or not.
The buildings that the REIT rents out will wear out, and eventually become worthless (or worse, have a negative value!) even under the best of maintenance conditions.
The fact that this process takes a lot of time to go over, and is not consistent across the board, makes many people disregard it in their valuations, but this is a constant force of gravity that is bringing down your earnings year after year.
Eventually everything decays.
Now, many people like to argue that while the building itself might eventually decay, the land (and its location) will not, and so the depreciation schedules don’t really matter as much since you will still own the land long after the buildings on it are gone.
This is an argument favored by new investors who don’t have a clear understanding of current GAAP principles regarding Real Estate.
The land value of the property is not depreciated.
While GAAP and IFRS have some differences in how they handle land on the balance sheet, the fact is that in both systems land is not something that can be depreciated.
The way this works is that upon acquiring the property, and based on comparable sales and other factors, the value of the land is determined, as are the value of the building and other intangibles that go along with it.
The value of the land is not depreciated, but the buildings (and other intangibles including rents in place) are in fact depreciated according to pre-determined schedules.
While these schedules will not be 100% accurate for every property (or indeed any property at all), on a statistical basis I find it very difficult to believe that there are meaningful inefficiencies and biases in the depreciation schedules.
As a general rule I believe that the schedules used are more or less an accurate representation of the statistical reality, and I personally don’t know enough about real estate accounting to think otherwise.
This post goes in depth on the exact accounting details, and I would heavily suggest you read it.
Of course, just because the depreciation schedules aren’t biased, doesn’t mean there are no other biases when it comes to real estate accounting…
But then what is that bias?
Real Estate Accounting Biases
There is a bias in GAAP that causes a systematic under-valuation of real estate assets.
We can see this due to the rules on how to determine impairments to the asset value of a real estate asset:
For example, let’s assume that in 2010 the undiscounted future net cash flows expected to be generated from a commercial property are $30,000,000 and the carrying value of the property is $50,000,000.
These projections imply that the real estate is impaired and it must be written down to fair value.
In order to calculate impairment, we will discount the cash flows to present value using an interest rate that is appropriate for the risks involved in generating the cash flows.
If the property appreciates in value in the future, however, no write-up to fair value is allowed under current GAAP.
In other words, the value of the property on the balance sheet is always the lower of either the cost basis of the property, or the fair value of the cash-flows it can generate.
Assets can be impaired, but cannot be re-evaluated to a higher value.
This means that non-cash income coming from asset value increases is not reflected in the balance sheet, which implies a systematic undervaluation of real estate assets on the balance sheet.
This puts a downward pressure on asset values, which meddles with the balance sheet and associated ratios.
After all, if you’re calculating Returns on Assets, or Returns on Equity when your asset value is not accurate, your ratio is going to be inaccurate as well.
I suspect this is why cash-flows are more often viewed as a more accurate method of valuation for REITs, not because depreciation doesn’t matter, but because they give investors a more accurate valuation of the underlying asset!
In other words, AFFO should not be seen as a measure of earnings, but as a proxy for an accurate balance sheet!
What this means for valuations
Unfortunately what this means is that I still have no good way to accurately value REITs.
To be clear, I’m not saying that there is no way to value REITs.
The thing is, when I am trying to value a company, I need the valuation method to have the following characteristics:
Be consistent - It should not change from company to company
Be cross-sectional - It should be applicable to every industry
Be financially Sound - It should make sense from a finance perspective
Be Simple - It should be easily calculated and understandable
Unfortunately REIT valuation always fails at at least some of those things.
The key problem here is the non-cash income that is not reflected in the income statement as a result of the fact that assets cannot be re-valued to a higher value.
This “income” is a key metric, but not one that can be found in a consistent manner, and indeed even if we were to find it, it would be inconsistent on a yearly basis.
I suspect that the best way to determine this value, or at least a proxy for this value, would be by looking at the location level data, and seeing what the rate of rent increases is established in the rental agreements.
This way, we would be able to get a general idea of the increased value of the existing real estate assets over the period of the leases.
Another thing to note is that REITs need to return substantially all capital to shareholders, expansion is usually achieved at the cost of dilution of existing shareholders.
Assuming the market trends towards being efficient, that means the bulk of the returns to shareholders will come from the dividends the company pays, combined with the non-dilutive portion of earnings increase.
In order words, your return will be the dividend plus the increased earnings coming from the scheduled rent increases in the contracts of the properties.
The issue I see here is that this is already supposed to be accounted for in the balance sheet!
For example, from Realty Incomes 10-K:
The majority of our leases are accounted for as operating leases.
Under this method, leases that have fixed and determinable rent increases are recognized on a straight-line basis over the lease term.
Any rental revenue contingent upon our client’s sales is recognized only after our client exceeds their sales breakpoint.
Rental increases based upon changes in the consumer price indexes are recognized only after the changes in the indexes have occurred and are then applied according to the lease agreements.
Contractually obligated rental revenue from our clients for recoverable real estate taxes and operating expenses are included in contractually obligated reimbursements by our clients, a component of rental revenue, in the period when such costs are incurred. Taxes and operating expenses paid directly by our clients are recorded on a net basis.
So as we can see, only certain increases are not accounted for already, namely Consumer price increases, and increases dependent on store sales.
Ultimately I don’t think this is the way to go when it comes to valuing the business.
Perhaps a better option would be to see the operating lease schedule in order to figure out the future revenues, and then go from there?
The problem with that is that the average lease duration will also affect that method, and in any case it must be compared with the debt repayment schedule as well in order to give any accurate assessment…
Summary
As a whole REIT valuation is a great big mess, and to be perfectly honest I am not confident enough to be able to value it in an accurate manner.
What does this mean for my second largest holding, Realty Income?
Honestly at this point I don’t know.
I get the feeling the company is doing just fine, despite trading at a very high valuation in terms of PE.
Part of the reason for this feeling of mine is due to the long lease duration, high rate tenant base, and the in-existence of any debt cliffs combined with minimum future annual rent payments that are more than enough to repay said long term debts.
If you have a better idea on how to value REITs, please let me know in the comments below.
This is really something I need to get a grip around, and I just haven’t been able to.
And of course, I’ll see you next time!