July 10, 2017 - 7:51pm EST by
2017 2018
Price: 22.60 EPS 0 0
Shares Out. (in M): 952 P/E 0 0
Market Cap (in $M): 21,515 P/FCF 0 0
Net Debt (in $M): 12,763 EBIT 0 0
TEV ($): 34,278 TEV/EBIT 0 0

Sign up for free guest access to view investment idea with a 45 days delay.



With many negative headlines about the “death of retail”, I figured value investment opportunities may present themselves. The problem I had was that I couldn’t determine which retail concepts would work and I’m no good at calling SSS comps. At the same time, it was also a tough call to decide which lower quality mall REITs that show the most upside would survive/exist in their current form. The catch was the higher quality mall REITs held up well and didn’t show much upside until recently (last ~ 12 months).

With even the “high quality”, “class A” mall REITs trading meaningfully lower, I believe GGP Inc. provides the best upside/downside with a significant catalyst and good shareholder alignment.

My investment thesis:

  1. GGP is a high quality retail REIT that transformed itself after getting in trouble during the financial crisis. The quality of its assets is apparent with GGP posting same store NOI growth in the 4-5% range and the ability to profitably redevelop department store boxes. Valuation looks attractive given my range of $20 - $30+.

  2. My analysis of the retail landscape suggests: a) US is over retailed and lower quality retail assets will face challenges, not likely Class A malls, and b) the other problem is not only online e-commerce (Amazon) but that the mix of retail category/format. Fortunately, GGP has demonstrated its ability to redevelop department stores and sell an interest in these redeveloped malls at very attractive cap rates.

  3. Brookfield Property Partners (BPY) is a 35% holder of GGP. It is likely Brookfield will use this opportunity to take GGP private, as they have done with various previously publicly traded companies they held in the past. In fact, GGP is the only one left. The November 2017 expiration date of Brookfield’s GGP warrants will likely serve as a catalyst for this. There are numerous reasons why Brookfield is highly incentivized to take GGP private. Fortunately, a very generous compensation package for GGP CEO Sandeep Mathrani highly incentivizes him to get non-Brookfield shareholders (and himself) a fair deal.

  4. It is usually tough to have an edge in larger cap names but I believe the opportunity exists because: a) general fear of death of retail, b) Goldman Sachs put a sell on GGP recently, c) None of the analysts covering GGP also cover BPY



GGP was originally founded by the Bucksbaum brothers in 1954 with a focus on shopping centers and went public in a 1972 IPO. Feeling the capital markets didn’t properly value GGP, the brothers sold off some assets and took GGP private in 1984. Over the next decade, the brothers realized that retail development projects were less plentiful and the next stage of growth had to come from acquisitions – meaning they need access to capital. The brothers took GGP public a second time in 1993 and grew its portfolio substantially through both mall construction and acquisitions – including the Homart Development Company (Sears development arm) in 1995 and Rouse Company in 2004. Perhaps driven by the second generation Bucksbaum family’s hubris, GGP became more comfortable financing acquisitions with debt – in fact the Rouse acquisition was essentially financed entirely by debt, bringing GGP’s debt to EBITDA ratio to ~ 18x.

Although GGP became the second largest US mall REIT (second to Simon Property Group), the 2008-2009 financial crisis essentially froze credit access for GGP, ultimately leading the company to file for bankruptcy protection in early 2009 after it was unable to repay $900 mln in debt secured by its Las Vegas properties. Operationally, it is important to note GGP’s portfolio was still > 90% occupied with NOI growth in the 2% range during the crisis. Its portfolio consisted of 200 malls with ~ 160 mln sq.ft. of GLA, a property management business and a Mastered Planned Communities (MPC) business.


Emergence From Bankruptcy

Heading into 2010, it became increasingly clear that GGP’s asset value exceeded its liabilities and that there was still value in GGP’s equity. As such, there was a competitive process where restructuring/recapitalization proposals were made. Without going into the details, it is interesting to note Simon Property Group submitted a total of four bids. In the end, the Brookfield/Fairholme/Pershing Square $6.3 bln recapitalization proposal was approved by the courts – including 7 year warrants issued to the consortium (as an aside, to be discussed later, it has now been 7 years and Brookfield is the only one holding the remaining warrants).


Source: GGP

Exiting bankruptcy, GGP named Sandeep Mathrani, the former executive of retail real estate at Vornado, as the new CEO. With a new focus on shedding itself of non-core assets and lower quality malls, GGP began focusing on high quality malls by: 1) spinning off Howard Hughes (the MPC business) in late 2010, 2) spinning off Rouse Properties (35 lower tier malls) in early 2012 – which was subsequently taken private by Brookfield in 2016, and 3) transferring 13 lower quality retail real estate assets to lenders (referred to as Special Consideration malls in GGP’s 2010 10-K).

Current GGP

Having sold off lower quality mall assets and done some redevelopment, GGP currently has 127 retail properties with 121 mln sq ft of GLA – of which 78 are considered class A malls and the rest are predominantly class B. 2016 NOI was $2.36 bln. 952 mln diluted shares outstanding. Net debt at $13 bln. Debt to EBITDA < 8x. In line tenants generate > $580/sq.ft. in retail sales.

US Retail Landscape – Death of Bricks & Mortar Retail?

Before we go onto valuing GGP, the biggest question to answer is how we handicap this “death of retail” risk and whether Amazon (and other online retailers) will take over all facets of retailing. I think the short answer is retail is not dead, and Amazon/e-commerce, while a threat, is serving more as a catalyst to the real underlying problem in the US – the oversupply of retail real estate (and the retail mix is also misaligned)

My discussion below makes references to the data obtained from the US census bureau on retail sales and e-commerce:

With data going back to 1998, we can see 1998 US total retail sales was $2,582 bln, while e-commerce retail sales was ~$5 bln – which also means retail sales ex-e-commerce was $2,577 bln. In 2016, total retail sales was $4,863 bln and e-commerce was $391 bln (AMZN being about 1/3 of that) – so retail sales ex-e-commerce was $4,472 bln.

So the 18 year CAGR for total retail sales = +3.6%, e-commerce = +27.4%, and retail ex-e-commerce = +3.1%

First observation is looking at historical data, there doesn’t seem to be a big problem. B&M retail sales growing at 3.1% is still pretty healthy and will support mall rents and growth. Looking forward, how I look at it is what kind of e-commerce growth rate and over what period would B&M retail sales stop growing even when total retail sales continue to grow at 3.6% on average? There are many different possibilities but the over/under here is something like 20% e-commerce growth next 5 years and 15% growth thereafter where B&M retail sales stop growing by 2026. So this oversimplified fairly pessimistic case B&M can continue to grow for ten years (sustaining retail rental rates). In reality, this scenario is probably too negative because:

  1. Actual e-commerce sales growth slowing to +15% in latest data

  2. Consensus sell-side estimates on AMZN has them growing top line by 21% in 2018, 19% in 2019 and 18% in 2020 – the North American portion would have less growth relative to international and whatever else Bezos have in mind (i.e. AWS and other new things). The other 2/3 of non AMZN e-commerce may not necessarily grow at or better than AMZN rates.

  3. Difficult to quantify factors including: a) this “omni channel” retail strategy taking place – which I think really means a B&M retailer also sells online and if a customer buys online allow them to return the product in a B&M store so you end up upselling/cross selling them for more, b) online retailers beginning to open B&M stores shifting some e-commerce retail sales back to B&M (i.e. Apple stores, Microsoft stores, Bonobos and even ironically Amazon)

So maybe things aren’t that bad and we have at least a decade or more in runway for B&M retail sales growth – why are retail/mall REITs under such pressure?

I think part of the answer is the US significantly overbuilt retail real estate in the past 40 years. The chart below nicely captures this dynamic – with US sales per capita at $14,614 and retail sq.ft. per capita at 24. While I don’t know what the “right” answer is, conservatively assuming the US needs to right size itself similar to Australia would imply ~ 17.5 sq.ft. per capita. Rightsizing to Canada’s levels would imply ~21 sq.ft. per capita. I’m going to guess 12.5% to 27% of US retail real estate sq. ft. may need to go. For this write-up, I’ll use 20% as a rough guess.

Source: GGP, ICSC

Compounding the problem of retail square footage oversupply is arguably the wrong mix of retail as shown below:
Source: GGP, ICSC


All in all, I think the oversupply issue boils down to: 1) some retail sq ft needs to be demolished/repurposed to something else and 2) some retail sq ft need to be redeveloped into some underrepresented categories (entertainment and retail, food and beverage, etc.). Although it is hard to prove, my guess is the higher quality retail assets can be profitably redeveloped while the lower quality ones will be out of business or repurposed for lower end uses (i.e. negative return on capital to repurpose).

Mental Model to Handicap GGP Risk – Commodity Cost Curve

As you may notice from some of my previous write-ups, I look at commodity businesses quite a bit and one of the mental models I use is the cost curve to determine the downside to a commodity, assuming there is a supply response to dropping prices as higher cost producers curtail/shut production.

Applying this thinking to GGP is a little different, the supply of retail real estate and its competitive position is not driven by cost by more by quality, in terms of location, rents, etc. However, in an oversupply retail real estate market, it would make sense for the lower quality properties to shut/repurpose first.

At a high level, there are 10.5 bln sq.ft. of retail real estate in the US. Interestingly, less than 10% of that is represented by malls at 894 mln sq.ft. (GGP’s bread and butter) which also coincidentally commands the highest average rent ($19.81 psf) and second lowest vacancies (4.3%). While some categorizations are down to semantics but practically speaking, most of the US retail sq ft is actually in smaller, regional strip malls. It would appear a lot of this sq ft will have to shut or be repurposed (with questionable incremental returns on capital) to other uses. In other words, there is easily 2 bln sq.ft. + of lower quality retail real estate that should go before GGP’s. Without a lot of incremental retail supply, GGP’s rental rates actually stand to go up if this actually transpires (same amount of B&M retail sales over a smaller retail foot print).


Source: JLL Retail Outlook


As a side note, obviously retail real estate developers got the message on oversupply and new supply coming on is minimal. In fact, consultants/brokers assume new retail space coming on to represent < 0.5% US retail square footage in the coming years.

I recognize the analysis above is a bit broad brushed so let’s drill down a bit and say we don’t know which category of retail real estate will adjust to the 20% oversupply. Instead we can assume every category needs to lose 20% of square footage. In that case, we know there are about 1,100 malls in the US and that the quality of malls are distributed as follows (we can argue about these grade rankings but it’s basically by sales per square foot and some qualitative factors):


Source: Taubman Investor Presentation


So if 20% of the malls needs to close, that would be about 220 malls. GGP disclosed that 78 of its 127 malls are class A and the rest are broadly B or higher. Going by quality, a 20% cut would be all the D malls (72), C- malls (89) and just over half of the C malls (59) would have to go.

Recognizing the argument that some C and D malls are the only game in town for smaller communities, I would say there is an incremental ~240 malls of lower quality as a “margin of safety” from GGP’s perspective.

What if there is a lack of/minimal rationalization of oversupply? Perhaps rents would adjust lower

The above scenario somewhat requires rational behaviour where the lower quality uneconomic malls shutdown or repurpose into something else. We can make the argument this does not happen but retail sales ex-commerce grows in the near-term still at a higher rate than new supply (say 2-3% vs. 0.5%), then a small amount of oversupply gets absorbed.

In this scenario, I would handicap it this way: Looking forward if retail real estate supply is over supplied by 20% but retail sales (ex e-commerce) net of new supply can grow at 2% AND if there is minimal retail real estate closures, then rents would have to adjust by say 18% lower (20%-2%). The logic is not perfect as higher quality malls would attract tenants and perhaps maintain rents but this helps illustrate what a downside scenario would look like.


GGP Valuation

Instead of going for being precisely wrong, I hope to be broadly correct in the valuation range for GGP. As such, I really simplified GGP’s valuation under the two scenarios identified: a) retail supply contracts and rents stabilize and b) retail supply remains oversupplied and rents adjust downward by ~18% across the board.

Under scenario A, I get about $2.3 bln in NOI (including non-consolidated properties). For cap rates, I look at some recent transactions GGP has done in the high 3’s cap and use 4% cap rate for the high end of the range and a 7% cap rate for the low end (this was approximately the cap rate the market valued GGP at coming out of bankruptcy, at a time when GGP’s portfolio had lower quality properties (i.e. Rouse) as well). So it’s pretty rudimentary but I get a range of $21.50-$47.75

Under scenario B, I take rents down 18% across the board which is a bit draconian (and we can assume some cost cutting, offsets). Rough numbers I get $1.9 bln in NOI for GGP. I would probably stop at a 6% cap using these assumptions as rent risk is essentially removed under such a negative rent rate adjustment.


Bottom line, I put a GGP valuation range between $20 and $35 and the stock is trading at $22.60 – a very compelling risk/reward.

As an aside, GGP had great success redeveloping its higher end properties. They got rid of Sears and redeveloped the Ala Moana in Hawaii and sold an interest in it at a 3.8% cap:$55-Billion-Following-Investment-by-Australian-Pension-Fund/169476

They downsized Macy’s and Dillard’s in the Vegas Fashion Show Mall and sold an interest in it at a 3.9% cap:$125bln.html


Brookfield – Highly Incentivized to Take Out GGP and minority holders have CEO on side

I am going to spend a bit of time covering why Brookfield would likely take out GGP and possibly soon. Here are the main points before I go through the detail and nuances:

  1. Brookfield Property Partners (BPY) incentivized to grow assets under management (management fee charged on capitalization). It has taken every formerly publicly traded holding private with the exception of GGP. Please see my original BPY writeup:

  2. Brookfield has GGP warrants which expire November 2017. Obviously they would exercise and it would cost them ~$600 mln. Possibly needing to raise funds to exercise warrants, this may be a good time for a go private transaction – no point in going in half pregnant

  3. Brookfield sees opportunity in redeveloping higher quality retail assets. It is in parallel looking to redevelop Macy’s stores etc. Specifically Brookfield Property Partners created new category of “opportunistic investments” just for things like this. Seems strange to duplicate efforts. GGP itself constrained by REIT status to spend significant capital to redevelop.

  4. GGP CEO Sandeep Mathrani has had a generous compensation/retention package. He has meaningful skin in the game and isn’t afraid to clash with Brookfield in the event of a go private.

  5. BPY itself uses IFRS accounting to fair value its holding in GGP irrespective of GGP stock price. BPY pegs GGP at > $30 per share.

BPY Incentives

As discussed in my previous BPY writeup, BPY gets 1.25% of growth in total capitalization as a management fee (not just equity, not per share). This creates a strong incentive to grow assets and at the very least to rarely reduce its asset base. In addition to this incentive, BPY probably thought it could narrow its NAV discount by taking its public holdings private so investors couldn’t as easily duplicate its asset base – as well BPY could reduce costs and allocate capital more efficiently between different entities (i.e. it’s not easy to cut GGP dividend to zero and use it all towards redeveloping real estate if it’s a public company).

These incentives largely played out as every public entity except GGP has been taken private since the BPY spinoff:

  1. Brookfield Office Properties BPO-US Taken Private Late 2013 – Tender offer valued at US$5 bln. 15% initial take-out premium. Another 5% bump later on.

  2. Songbird Estates SBD-LN (Canary Wharf) Taken Private Late 2014 – 34% initial premium, another 20% bump later. Transaction valued at ~ US$4.1 bln

  3. Rouse RSE-US Taken Private Early 2016 – 26% initial premium, another 7% bump subsequently. Transaction valued at US$2.6 bln.


Granted, the pushback to a GGP take-out is that it will probably cost in the range of $20 bln and that Brookfield actually reduced its stake not so long ago. My counter to that is:

  1. Brookfield technically reduced its GGP stake but it distributed GGP shares to its LPs who invested alongside Brookfield during the restructuring. Brookfield likely had to do this to realize the carried interest for generating returns for its LPs. Interesting they chose to distribute shares that would leave an overhang instead of blowing it out.

  2. Filings show Abu Dhabi Investment Authority and Future Fund (Australian Sovereign Wealth Fund) together still holding 5% of GGP. BPY can likely structure a take private with the participation of these two SWFs amongst others, much like how BPY structured the Songbird Estates takeout with the Qatari’s.


GGP Warrants – November 2017 expiry

This one is pretty self-explanatory and should serve as a good near-term catalyst for Brookfield to act. It is interesting that GGP bought back warrants held by Fairholme and Blackstone and Brookfield bought Pershing Square’s warrants (after some agitation). Brookfield is the only one holding the warrants until the end – this is a strong signal of their intentions.


Brookfield owns 73,930,000 warrants (the “Warrants”) to purchase common stock of GGP with an initial weighted average exercise price of $10.70. Each Warrant was fully vested upon issuance, has a term of seven years and expires on November 9, 2017. Below is a summary of Warrants that were originally issued and are still outstanding.


Warrant Holder


Number of Warrants



Exercise Price

Brookfield - A








Brookfield - B














The exercise prices of the Warrants are subject to adjustment for future dividends, stock dividends, distribution of assets, stock splits or reverse splits of our common stock or certain other events. In accordance with the agreement, these calculations adjust both the exercise price and the number of shares issuable for the 73,930,000 Warrants. During 2015 and 2016, the number of shares issuable upon exercise of the outstanding Warrants changed as follows:






Exercise Price

Record Date


Issuable Shares


Brookfield - A


Brookfield - B

April 15, 2015












July 15, 2015










October 15, 2015










December 15, 2015










April 15, 2016










July 15, 2016










October 14, 2016










December 15, 2016










December 27, 2016












Brookfield has the option for 57,500,000 Warrants to either full share settle (i.e. deliver cash for the exercise price of the Warrants in the amount of approximately $618 million in exchange for approximately 72.5 million shares of common stock) or net share settle. The remaining 16,430,000 Warrants owned or managed by Brookfield must be net share settled. As of December 31, 2016, the Warrants are exercisable into approximately 62 million common shares of the Company, at a weighted-average exercise price of approximately $8.48 per share. Due to their ownership of Warrants, Brookfield’s potential ownership of the Company may change as a result of payments of dividends and changes in our stock price.


Brookfield Redevelopment Activities

It seems pretty clear that Brookfield does not think retail is dead (at least for higher quality retail real estate). In fact, Brookfield itself separately partnered with Macy to redevelop about 50 stores.

This seems a bit strange considering GGP itself mentioned in its 3Q16 conference call they were acquiring 5 Macy boxes to redevelop. Again, GGP as a public entity and a REIT is restricted by its need to pay out dividends to maintain its preferential tax status. It would make more sense with a private GGP to redevelop Macy boxes, and at the same time allow BPY to charge a higher management fee and more optimally allocate capital without such constraints.

In fact, in a recent change in messaging, a late 2016 BPY presentation shows they are now allocating ~ 20% of their capital to “opportunistic investments”: