2023-11-09 06:30:00 ET
Summary
- Safehold's value lies in its long-term, growing cash flows from ground leases and its ability to expand its portfolio.
- Today, their cash flows are challenged, but not in a way that threatens the long-term story.
- Buy SAFE as a strong play on interest rate declines, but not for the dividend, which is not secure.
The big picture on Safehold (SAFE) involves the very-long-term, growing cash flows from their ground leases, their ability to grow their portfolio, and the terminal value of their leases.
I've published several articles on these aspects, most recently these two . Since ground leases are like supercharged long bonds, the change in the net present value ("NPV") of the Safehold ground leases is larger than the change in long interest rates. So that NPV has dropped substantially as interest rates have moved up since early 2022.
My argument about all that has been that the market has never recognized the value of Safehold assets, and on top of that over-reacted to recent interest-rate increases. Today, with interest rates so high and sentiment on this name so low, SAFE is a focused bet on interest-rate decreases.
Time will tell what interest rates do and how the market responds. Those are not my topic today.
The immediate Safehold cash flows have attracted quite a bit of interest. In my view this misses the point, which is the long-term story. But I did look at those cash flows in my article following Q2 earnings.
In that article, I projected the rest of the year and near-term years. This was tricky because it was not yet clear what aftereffects of their merger at the end of Q1 might be.
The Q3 results give me a chance to see whether my projections missed anything significant. They also let me refine my view of dividend security.
Here is the story.
Removing GAAP
My goal in working with the Safehold financial statements has been to find and then project forward the cash earnings that result from the main components. In preparing the present article I dug deeper, going through every line in the Statements of Operations and of Cash Flows and looking for what the 10-Q had to say about it.
In the process, alas, I found some formula errors in my previous spreadsheet. For the present article, after sorting all of it out I checked every formula in every cell. Even so, my qualitative conclusions for 2023 did not change from those offered in my previous article.
Going forward the Safehold cash flows should be simple:
- Cash rents on their leases
- Fee and other income (much smaller)
- Interest expenses
- General & Administrative ("G&A") expenses
- Dividends
Of course, there will be some noise, but those should be the main items. We can note that the fees for managing Star Holdings (STHO) are included as part of GAAP revenues.
If you look at my projection of the Income Statement for 2023, there are two other items above $10M for the year. These are "earnings from equity method investments" and a non-cash impairment.
The $145M impairment impacts Net Income but not GAAP revenues. There is no interest here in doing philosophy about impairments.
But there is still the challenge of getting from the GAAP numbers shown on the Income Statement to the cash numbers we care about. Fortunately one has the cash flow statements to help.
There are four non-cash items above $5M related to revenues. These are
- Stock-based compensation
- Deferred operating lease income (essentially straight-line rent)
- Non-cash interest income from sales-type leases
- Non-cash management fees
There is also non-cash interest expense projected at $13M.
Similarly, about $20M of small, typically non-cash items are expensed on the Income Statement. The several small items added back in as part of finding Cash from Operations ("CfO") roughly balance these items.
We ignore both sets of small items in seeking cash numbers here but note that the balance is better in some years than in others. These items are responsible for much of the difference between CfOex (CfO ex changes in working capital) and my calculation of simple cash earnings.
There also is the impact of so-called equity method investments. The net profits from these contribute to GAAP Net Income but not GAAP revenues.
On the Statements of Cash Flows the entire profits are subtracted out and then the portion of them distributed as cash to Safehold are added back in. I added this item as part of cash revenues.
Here is my revised projection of the primary cash flows for 2023. For nearly all financial quantities, Q2 and Q3 produced similar values. This markedly improved the accuracy of my projection of the full year, but overall the changes mostly offset one another. Here is my current projection.
The topmost sector, shaded green, subtracts the big, non-cash items from GAAP revenues to get the inferred cash revenues. The non-cash part runs nearly a third of the total.
There was a drop in the projection of non-cash adjustments to revenues, and specifically the adjustments to lease income to take out straight-lining and similar adjustments. Since the amounts in Q2 and Q3 were comparable, Q4 should be the same and there is no longer a need to project these items as a fraction of various revenues.
The next sector, shaded orange, shows the cash interest expense. It also finds this to be about 70% of cash earnings, as one expects from the Safehold business model. (This very atypical value reflects the fact that Safehold is a very atypical REIT.)
The section shaded pink shows the cash G&A, found after subtracting stock-based compensation. This is coming in a bit lower than the $40M/yr run rate discussed in the Q2 earnings call.
The section shaded blue shows the resulting simple cash earnings, CfOex, and the dividends paid. The dividends are not covered by cash flows. This was also my previous conclusion. More discussion on that is below.
It is also, unfortunately, worth noting that NAREIT FFO (Funds From Operations) will not have subtracted those non-cash adjustments to revenues, which are huge. Any discussions of FFO payout ratios or FFO multiples for this REIT are deeply misguided.
Interest Expenses
It is worth a brief discussion of interest expenses since various analysts have written as though the floating-rate debt of Safehold is a major problem. One is hampered by the opacity inherent in the way derivatives are portrayed in the SEC filings.
Fortunately, CFO Brett Asnas provided a clear explanation on the Q3 earnings call:
Similar to many borrowers, we have felt the effect of the rapid pace of rate increases. Over the last year and a half, we have mitigated the impact by putting in place $500 million floating to fixed swaps … [it has been made clear elsewhere that the resulting interest rate is 3%.]
We have $400 million of long-term treasury locks at a weighted average rate of approximately 3.4%. At current rates, these are approximately $75 million in the money. … While these hedges protect us through next year, they can be unwound for cash at any point. As we look to term out revolver borrowings with long-term debt, we would unwind the hedges and attach the gain to the debt, lowering the effective economic rate we pay.
Here is my breakdown of the debt and interest expenses based on the Q3 reports:
The rows shaded blue show the distribution of the total debt by type. There is fixed debt, with a 23-year, weighted-average maturity. Then after the $500M of debt covered by the swaps, there is $603M left. (In addition, Safehold used funds from an equity raise during Q3 to pay down about $150M of floating-rate debt.)
The rows shaded green show the breakdown of the interest expenses. About two-thirds of them are on the fixed debt, with another 9% on the swapped debt. Estimating the rate on the floating debt at 6%, one ends up with the estimate that various other costs that flow through to the interest expense line total about $3M.
The point worth making here is that the increase in interest expenses has been driven almost entirely by the increase in total debt that accompanied growth of the portfolio in 2022. Had we not seen the massive increase in interest rates over the past two years, the interest expenses would be less than $10M smaller out of more than $160M. The common narrative that the fact so much debt was floating is the main factor creating challenges is not correct.
Looking Ahead
To project future years from here, I assumed no growth of the portfolio, just to see what was implied. We return to growth in a bit.
The bulk of revenues is rents from ground leases, which have escalators that average 2%. But the fees from managing STHO are significant, so I projected these separately. I lumped the other, small, income items in with lease revenues.
The fees from STHO run $25M in year 1 from March 31, 2023. Then they are $15M, $10M, $5M in the next three years (which is as far as I went).
The net result is that the sum of cash revenues from the existing portfolio and the fees from STHO will be nearly constant over those years. I also kept G&A costs at $40M, although they will decrease somewhat as the STHO properties are sold. Interest costs might drop but we saw above that this will not have a big impact.
The model projected the Q3 interest costs forward. Then we find this:
If the assumptions used here were proven correct, then going forward cash earnings would fall short of the current dividend payments by about $30M per year. In the short run this is no big deal, considering that this is less than 1% of the current debt.
For comparison, Federal Realty Trust (FRT) elected to sustain their dividend across the pandemic, generating cashflow losses (CfO less capex less dividends) above 5% of their debt in 2020 and 2021. [Safehold has no capex.] Federal took almost no criticism for this, even though their long-term cash flows are less secure than those of Safehold.
I like Federal a lot (article here ) and would have made the decision they did. But the difference in the tone of the criticisms does not strike me as objectively justified.
Returning to Safehold, the board might decide at any time to decrease the dividend so that it was well covered. This would be a big drop. No one should buy SAFE for the current dividend.
Why Are We Here?
The idea of getting into this situation had three parts, all from the viewpoint of late 2021. First, Safehold had grown large enough that they needed to internalize management. Second, they were growing the portfolio at well north of $1B per year (it reached $2B for the year ending in mid-2022) and had every reason to expect that to continue.
Third, their approach (a common one) was to fill up half their revolver and then place the debt. Unfortunately, the revolver was half full when the world fell apart.
Had Safehold management known the future, one would think they would have made different plans. Even so, if you look today at the long-term value of the contractual cash flows produced by their ground leases, things look fine (as I discussed here ).
But cash flows are challenged.
But What About Growth?
The current difficult cash flows will continue until the portfolio grows. The net cash earnings from new leases are about a third of the total lease income. So in round numbers Safehold needs $100M of new lease revenues to close the cash-flow gap.
We don't know where the real estate markets will settle. If the new stable rate becomes 5% on a ground lease, then getting that $100M would take $2B of portfolio growth.
That is not a lot of growth and has been done in a year before. More generally, I am sure Safehold management expects to put the cashflows in good order within a small number of years after the new real-estate markets define themselves.
One of my investing friends was grumpy after the Safehold earnings release. What bothered him was precisely the lack of portfolio growth - the tiny amount of new ground lease signings.
Paraphrasing him: "You guys are charging me and other owners $40M a year plus stock-based compensation with the stated purpose of growing the portfolio. So far you have been wasting our money. Get busy or shrink those costs."
I know he sounds like an impatient owner of an NFL football team. And the capital and real estate markets are still difficult, as CEO Jay Sugarman highlighted in the Q3 earnings call.
Even so, skeptics will suppose that the markets might find a new state where Safehold is closed out of doing much new business. This seems impossible to me, and management will jump up and down on their desks about why this will not happen. But that will not sway the skeptic.
In that eventuality, Safehold ought to shut down the attempt to grow, shed a lot of people, and pull G&A costs down to a level sensible for a steady state company running a ground-lease portfolio.
Takeaways
Once one unpacks the real cash flows from GAAP, one can see that the cashflow challenges are real but are in no way material to the future of Safehold. With their high credit ratings, issues at the $30M level just are not significant.
SAFE should move very strongly upward when interest rates drop, and that is the reason to buy the stock today. SAFE should not be bought for the dividend.
Longer-term, the terminal values of their leases are worth attending to. And there should be a growth story.
That growth is central to improving the cash flows and to improvements in stock price beyond the coming bump when interest rates drop. But as I told my friend, it could yet be a few quarters before we see much.
For further details see:
A Check On The Safehold Cash Flows After Q3