Description
Several notes:
- I am not bullish on REITs or other fixed-income-like products so this idea should be thought of as a long-short with VNQ (or your favorite REIT short) as the short.
- This was written up as a short idea by nha several months ago. The comment thread is instructive. The stock is down 17% since nha's write-up (4.87 to 4.06) while VNQ (including dividend) is down 11%.
- While the stock is down versus REIT index, the company has done a good job of executing versus the business plan and achieving at least some scale so that the downside is now much closer to the current price, and the upside remains in place and a quarter closer.
Long Synposis:
I think the short argument here rests on two points: (1) The arbitrage between the public and private markets does not really exist. It is not so simple as comparing NOI yields to REIT equity WACC. However, the simplicity of these deals: single-tenant, high-credit quality, long-term, triple-net leases (and the obvious alternative uses if tenant defaults or walks away after term has ended) seems to offer an obvious response. This argument reminds me of the EMH economists who walk by a $100 bill on the street . . .
(2) The more serious argument is that while this relationship may exist for a short window, it is going to evaporate away quickly. Thus, the core of the conflict between the longs and shorts in this name is how quickly GPT can scale and whether they can then raise a secondary at an inflated price ($5-6). The company is exploiting a very simple "arbitrage" between the private leasing market and the public REIT market. In broad strokes, GPT is buying properties that are yielding 9% unlevered NOI with no incremental overhead while funding that incremental 9% NOI at less than 6% WACC. If the arbitrage had closed in Q1 of this year, GPT as a standalone company would not have been worth much more than TBV as they would have had a lot of unused cash and would not have achieved the scale necessary to overcome their overhead fixed costs.
NHA's short argument 4 months ago is consistent with this thesis. At the time of nha's write-up, I had thought that GPT's stock price had gotten a bit ahead of the story and while I was cautiously optimistic that the arbitrage would last longer, the $5 stock price had built into it that the arb would persist throughout 2013. However, four months later, we know that the arb window did remain open and GPT capitalized on the opportunity--doing well over $100mm in purchases at 9+% NOI rates. At this point, even if GPT never did another deal, the stock is now worth $4-ish per share. And if GPT completes another $100mm in purchases (using up its existing capital), then it is worth $5 per share. Of course, the upside is capped only by the willingness of the seconday market and the continuation of the private-public "arbitrage". The upside story has not changed in GPT over the past few months, but the downside scenario (deal window closes immediately) is not much lower than the existing share price.
To fully scale with existing equity capital, GPT needs to buy $450mm (including equity portion of JVs) or real estate with 9+% NOI yield (adjusted for inflation-adjusted rents). As of 7/29/13, they have $350mm (purchase price) in properties, and they closed on over $100mm in the past 3 months. The deal-making is lumpy, but they should be fully invested into current equity capital in the first month of Q4.
I would add a third bear thesis (that I have not heard from others), and that is that the asset management business dies after the existing contracts roll. Management seems to think they have a large moat with existing clients (only four) and the the business will continue, but this is a significant risk.
Valuation Methods
I do not have any magical valuation methods, and I would refer to nha's write-up for three basic methods. I do have thoughts about each.
First, I agree with nha's criticism of considering P/B ratios blindly across similar REITs. I do not think it is a worthless exercise as book value may mean a lot more on a long-term lease than fair market value of property, but this requires far more nuance (current FMV; length of lease; cap rates; etc.) than simply lining up P/B ratios as GPT's management does.
Second, the 7% unlevered cap rate (of core NOI--before fixed-cost SG&A) offered by the company is an attempt by the company to argue for a scaled valuation metric before scale has been achieved. It is not an unfair argument, but bulls must acknowledge that it will significantly over-value the company if the arb window closes and scale is not achieved. In order to use a 7% unlevered cap rate, the $13mm of fixed overhead costs must be spread over more than the existing properties (and even more than the $450mm in the event GPT maxes out without raising new equity). The 7% unlevered cap valaution method becomes appropriate around $650mm of total property -- which means one secondary offering. Interestingly, at the time this method becomes the correct method (if it is ever the correct method), it will undervalue the current equity because it will not take into account raising additional equity at a premium to book.
So I think scenario-based FFO analysis offers the best perspective at this time. I would offer three scenarios:
I am using a 5.0% WACC of debt and pfd, and I am using a 7% cap rate on FFO to equity. I think the former is quite high based on the BofA JV non-recourse financing of 4.6% and more recent debt financing. In the short run, the preferred will drive it up slightly over 5%, but once they take out the preferred, the WACC (non-common) should be less than 5%.
(1) Arbitrage closes immediately and no more deals done. ($0.17 FFO per share + $1.60 of cash-like assets that do not contribute to NOI or AM) --> $4.00
NOI from property: $32.7 mm
AM Contribution: $4.0 mm (note that this is lower than nha had because even though in a REIT, this business is taxed)
SG&A: $13.1mm
Interest/Pfd Dvd: $13.7mm (5% of $273mm (includes JV debt and unpaid pfd dvds ($35mm)))
FFO: $9.9mm
----------
Cash at Parent: $56mm
Cash at JV (share): $4mm
CDO: $8mm
CDO Advance: $10mm
KBS Promote: $12mm -- only $7.8mm recognized thus far, but mgmt confident full incentive fee will be earned
Taxes on remaning KBS Promote: ($1.4mm)
Assets for Sale in JV (share): $10mm
Garrison Promote: $0
(2) Current equity maxed out but no secondary. financing: $50mm of cash + $85mm of borrow - $35mm of pfd true-up. ($0.28 FFO per share + $0.75 of cash-like assets) --> $4.73
NOI from property: $42mm
AM Contribution: $4.0mm
SG&A: $13.1mm
Interest/Pfd Dvd: 16.2mm (5% of $323mm)
FFO: $16.7mm
(3) $100mm secondary at $5 per share done in Q4 2013 and fully invested in $250mm of property. (0.40 FFO per share + $0.75 of cash-like assets) --> $6.46
NOI from property: $65mm
AM Contribution: $4.0mm
SG&A: $13.1mm
Interest/Pfd Dvd: $23.7mm (5% of $473mm)
FFO: $32.2mm --> diluted by 33% --> $24.2mm
Note: I should discount some of this scenario 3 as this is on average 3 quarte rs away. But flipside is that some of the cash-like assets should turn into cash and be invested as well.
I believe that at minimum scenario 2 will be achieved and proven out within the next 2 months. As well, with a dividend and secondary announcement, some of scenario 3 will also be priced in the next 2-3 months.
-------------------
Accounting Confusion
There are two accounting issues that may be hiding the value:
- From its former life, GPT had consolidated CDOs on its balance sheet. It was a mess until they cleaned it up in Q1 2013. However, a chunk of CDO equity remains on the books. Management insists that is is conservatively valued, but clearly it is highly levered to underlying debt.
- More importantly, the JV accounting has produced quite a mess, but this will be going away starting in Q3 2013. When GPT bought the pool of properties known as BofA JV, they bought with it a defeasance pool (i.e. they were buying pledged Treasury securities that were offsetting the loans on the pool of properties) for $1 (i.e. add a quarter and you can get a candy bar) even though the Treasury securities actually produced about $4mm of cash flow too much. For some reason (and I cannot figure it out -- please let me know if you know), defeasance pool accounting has the pledged securities on the balance sheet as sum of cash flows (undiscounted) while the loans go on at par. And then over time, the interest expense on the loans is included on the income statement but the offsetting interest from the pledged securities does not because they are on at undiscounted rate. Similarly, when they sold the defeasance pool in Q2 2013, they then took a loss for the same reason. So this defeasance pool, which produced $4mm in simple gains (of which half go to GPT), produced quarterly losses and the one-time loss on sale. It should have had no effect until it was sold and then it should have produced a $2mm gain for GPT.
The JV is not consolidated on the balance sheet and shows up only as a $50mm JV Equity Investment.
The BofA JV economics are instructive and account for almost half of the current property held by GPT. The JV owns $290mm of purchased property (with NOI -- including asset management fees specific to portfolio -- of over 9%) funded by a $200mm of a non-recourse 4.6% loan. So we have a debt-to-equity ratio of about 2:1 and a low borrow cost. The importance of this capital structure is that it gives us some idea of what GPT's overall capital structure will look like once they clean up the preferred.
-------------------
Management
Management has laid out its plan and executed very closely. They return emails and respond to questions. They are one of the more straightforward teams I have seen. They openly state that the business plan entails instituting a healthy dividend and issuing secondary shares so that they can exploit this opportunity to the max.
--------------------
Risks
- Public/Private Arb gets closed out
- Asset Management business ceases after current contracts roll
- General REIT forces (neutralized with VNQ short)
---------------------------
Catalysts
There are three catalysts in the coming quarter:
- Q3 numbers will be cleaned of the Defeasance Pool accounting, and they will include the $100+mm of properties bought at the end of Q2 and beginning of Q3. Thus, we will start to see the scale effect.
- The company will make further progress on buying properties and should be fully invested (with current equity base) and so will be prepared to pay a dividend to common as well as pay the accrued dividends to the preferred. Management has made clear that they intend to raise more equity and so I expect a very healthy dividend in the $0.30 range. I expect a dividend announcement as soon as management has deals in place to use up the rest of capital (about $100mm in additional purchases).
- The company will announce a secondary offering as soon as it announces a dividend. Contrary to the norm, I think this stock rallies on a secondary as the incremental return on capital is much higher than the average return due to the fixed-cost overhead.
I do not hold a position of employment, directorship, or consultancy with the issuer.
I and/or others I advise hold a material investment in the issuer's securities.
Catalyst
Catalysts
There are three catalysts in the coming quarter:
- Q3 numbers will be cleaned of the Defeasance Pool accounting, and they will include the $100+mm of properties bought at the end of Q2 and beginning of Q3. Thus, we will start to see the scale effect.
- The company will make further progress on buying properties and should be fully invested (with current equity base) and so will be prepared to pay a dividend to common as well as pay the accrued dividends to the preferred. Management has made clear that they intend to raise more equity and so I expect a very healthy dividend in the $0.25 range. I expect a dividend announcement as soon as management has deals in place to use up the rest of capital (about $100mm in additional purchases).
- The company will announce a secondary offering as soon as it announces a dividend. Contrary to the norm, I think this stock rallies on a secondary as the incremental return on capital is much higher than the average return due to the fixed-cost overhead.