Description

I am recommending a long position in Prologis, as I believe today’s price (~30% off the April 2022 peak) represents a GARP-y entry point into a $100b+ market cap, ultra-blue chip, proven secular compounder in a market with many years of well above-GDP organic growth still ahead.

Executive Summary

Prologis is the largest US-listed REIT by market cap. Its primary business is owning and operating US industrial properties: 85-90% of revenue/earnings/FFO come from rental operations, ~85% of which comes from the US. The remaining 10-15% of revenue/earnings/FFO are generated by an investment management business with ~$60b of third-party AUM through which PLD manages industrial real estate funds invested in properties based largely outside the US. To align interests with investors, PLD holds 15-50% interests in each of its funds as unconsolidated JVs. Finally, PLD also builds new industrial properties to either contribute to its managed funds or own long term – to feed this pipeline, it owns an enormous land bank for future development at a total cost basis of $4.2 billion with the ability to support estimated future investment of $34.2 billion. 

Around two dozen sell-side analysts cover Prologis and not one of them has a sell rating. Its market cap is 2x larger than all other industrial REITs combined. It has beat or matched FFO/sh estimates every quarter since Q3 2013.  It trades at ~23x 23E Core FFO/sh and a ~4.0% implied cap rate on 23E NOI. 

A typical VICer’s first reaction to being pitched PLD as a long would probably go something like: “At 23x FFO / 4.0% cap rate vs ~4% 10y this is expensive. And isn’t their biggest tenant Amazon? There’s that guy on here who every time AMZN posts earnings claims their ecommerce business was purely a ZIRP phenomenon and will never make money… I’m not sure if that’s true, but it’s clear that e-commerce was a COVID beneficiary – industrial real estate must be too. These are just warehouses after all – we probably built way too many these last few years. If industrial fundamentals haven’t reverted to the mean yet, it’s only a matter of time before they do. PLD is clearly not a long for me and might even be a short.” This would seem to jive with a recent Seeking Alpha article called, “Prologis: No Alpha Here,” which recommends a ‘hold’ with an expectation of just ~8% annualized returns over the next 3 years.

I spend most of my time doing private market real estate investing – when I try to explain to my hedge fund friends why industrial is by far my favorite asset class right now, their eyes glaze over and they say something similar to the above. This writeup is inspired by those interactions as well as a comment that VIC member Blaueskobalt made on 1/21/23 on the BXP thread:

“Compare BXP (7.5% cap) to PLD (4.0% cap) & EQR (6.0% cap) - this suggests BXP is a good deal cheaper, but, once you mark each portfolio's rent to market and adjust for through-the-cycle CapEx burden, they are all three trading at a 5-5.5% unlevered cash flow yields.  Given all three trade at roughly the same price, PLD > EQR > BXP is a no-brainer, given the secular trends.”

I think this comment concisely hits every point necessary for the PLD long thesis. In particular: 

1. Reasonable Valuation: Given the embedded rent mark-to-market and low CapEx/re-leasing burden, PLD’s valuation today is less demanding than it appears at first glance. Underwriting only slightly above-GDP rent growth still results in compounder-type returns.

2. Strong Supply/Demand Fundamentals: Demand for warehouse space will continue to be a secular growth story due to e-commerce growth. Additionally, it is difficult to expand supply of warehouse space and will remain that way for the foreseeable future – this makes industrial real estate the constraining input in the e-commerce supply chain, and positions PLD to capture the super-economic profits generated by e-commerce.

PLD was last written up on VIC all the way back in 2008 by agape1095 as a special situation short – at that time 80% of PLD’s revenues came from new developments being contributed to the managed funds at artificially inflated prices, which the author described as a ‘house of cards’. This thesis was dead-on – the stock traded down ~75% from the mid $40’s to the low teens when the financial crisis hit a few months later, far underperforming the S&P 500’s ~33% drawdown over the same period. Management was replaced shortly after, and over the past 15 years PLD has refocused on its core business of owning and operating industrial real estate. 

PLD’s returns over the past 15 years vs SPY and VNQ (Vanguard REIT Index) can be seen in the chart below. If you had bought PLD on 5/6/2009 when the author recommended exiting the short (“the new management team is on the right track”), you would have realized an IRR (assuming no taxes and dividends reinvested as received) of 17-22%, depending if you had exited into the post-COVID highs (S&P Peak – 12/29/21, PLD Peak – 4/28/22) or today (apologies but the “today” is actually 3/3/23 – I couldn’t get my BBG excel sheet to refresh and had to post this by midnight). This would have outperformed SPY by 400-700 bps and VNQ by 700-800 bps. 

In fact, even if you had bought the day the short writeup was posted in Aug 2008 – right before a 75% drawdown – the secular growth tailwinds for the next 15 years would have bailed you out, and you would have still outperformed VNQ by 450-550 bps and slightly outperformed or matched the S&P. 

Driving these returns is ~11% FFO/sh CAGR from 2011 to present – FFO multiple hovered around 20x for most of that timeframe. According to PLD’s research, national industrial rents grew at around 5%/yr over the same timeframe. 

Here is some conservative back-of-envelope math for equity returns on a go-forward basis (will go into more detail in part 1 below): With rents fully marked-to-market, today’s price gives you Prologis at a ~5.0% cap rate / ~15x FFO multiple. Assuming rents grow 5% annually for the next ~20 years (2.6x total rent increase over 20 years), unlevered total return is 10.0%. Layer in 20% leverage at 3.0% cost of debt (current debt has wavg. interest rate of 2.5% and wavg. maturity of 10 years) and levered equity return goes to ~12%.

This 12% equity return gives no credit for either the development pipeline or G&A efficiencies at scale. I view the development pipeline in particular as a major driver of future returns: not only can PLD invest $30b+ at a yield-on-cost that’s accretive to its 5.0% implied cap rate (on product that will benefit from the same rent growth tailwind), but in many markets PLD effectively controls the entire future supply pipeline.

Additionally, the supply/demand dynamics that generated ~18% and ~26% rent growth in 2021 and 2022 show no signs of abating – rent growth in 2023 should hit double digits again, and with minimal new development starts in 2023 I think double digit rent growth for the next few years is very achievable. Adding all this together I think realized IRR for PLD should be in the mid to high teens. 

Part 1: Reasonable Valuation when Accounting for Lease Mark-to-Market

When signing a commercial lease, the tenant locks in a rental rate for a long period. Many times the rental rate is flat for the entire lease term, and it’s very rare for a lease to have greater than low single-digit annual rate escalations on average. In Prologis’ case, rents over the past two years have grown in the high teens / low 20’s annually – so most of Prologis’ leases signed before 2021 are at rates significantly below today’s market rents. 

I view underwriting this full mark-to-market as very safe. Industrial rents would have to decline dramatically for anything less to be realized – this hasn’t happened since the GFC. In fact rents today are still growing at a double digit rate. 

Prologis signed 174mm SF of new leases in 2022 (18% of total portfolio) at a 51% net effective premium and 32% cash premium to expiring rents. I view the fact that net effective rent change is so much greater than cash as a big positive for PLD – this means renewal/new-lease tenants are shouldering more of the CapEx burden and requiring less concessions than they did for first-gen leases – you would see the exact opposite of this phenomenon for most other property types, especially office. 

It is tricky to estimate exactly how much lease MTM to underwrite. Page 20 of PLD’s Nov 2022 NAREIT presentation shows $4.6b of in-place NOI being marked to $7.0b of “In-Place NOI Plus Incremental NOI from LMTM”, based on 62% estimated MTM. I am inclined to believe the $7.0b number is conservative because:

1. $2.4b on $4.6b is only 52% NOI increase – this reflects that most of PLD’s leases are NNN and expense recoveries don’t have MTM.

2. Management’s rent growth estimates have been conservative in the past (and they acknowledge this)

3. The 2022 10K mentions 67% net effective LMTM – up 5 points from 62% in the NAREIT presentation just 3 months earlier (and despite another ~5% of the total rent roll marking to market over that period)

Below is valuation math getting to a 4.0% in-place and 5.0% fully MTM cap rate. Note I’m giving no credit to strategic capital promote income. Adding the same $2.4b MTM cash flow to 2022 Core FFO results in the 15x MTM FFO multiple.

Part 2: Strong Supply/Demand Fundamentals

US E-commerce as a percentage of overall retail sales has steadily grown over the past decade. Now that the COVID pull-forward has worn off, US e-commerce penetration sits at about 15%. 

There are countries such as South Korea where ecommerce is already 30%+ of retail sales and still growing. Various forecasts project that US e-commerce penetration will rise to 22% by 2027.  I honestly don’t know who makes these projections or how they make them. What I am very certain of is that younger people are more comfortable shopping online than older ones. Even if ecommerce penetration at a cohort level stays flat or declines, ecommerce sales should continue to grow as the population ages.

E-commerce companies typically require 2x to 4x the amount of warehouse space as traditional bricks & mortar retailers. While e-commerce growth been responsible for driving industrial vacancy to record lows, it’s worth noting that Prologis’ leasing is fairly broad-based:

Today’s cities were not built with ecommerce or same-day delivery in mind. The Power Broker by Robert Caro (great book I would highly recommend) recounts how Robert Moses, the most powerful man in NY politics for decades, could easily have built one-stop transit access from Manhattan to the NYC-area airports in the 1950s – instead he chose to build highways with no mass transit. Today, the eminent domain cost alone of creating such transit is mind-bogglingly cost-prohibitive, and New Yorkers will forever have to either schlep their luggage between 2-3 different transit systems or sit in traffic to get to the airport. Although industrial real estate is developed privately, I think of the supply constraints on industrial real estate in much the same way. 

Industrial still commands by far the lowest rent per square foot of any property type. Prologis’ average rent PSF for its owned & managed portfolio is just ~$7 PSF. A relatively cheap 1000 SF apartment at $1,500/mo rent costs more than 2x more at ~$18 PSF. It would be rare in my mind to see a retail rent below $10 PSF or an office rent below $20 PSF. Also, with very few exceptions, warehouses can only be built to one story, and they require significant amounts of land be set aside for non-income producing parking and loading space. All other property types can be built to greater density with less of a loss factor and thus generate much more income-producing square footage on a given parcel of land. 

Speaking very, very broadly, this means that industrial real estate is effectively priced out of development sites that work for just about any other type of real estate. Even at the all-time tight 3-handle private market cap rates, many infill warehouses were still trading below replacement cost given the significant value of the land beneath them. As such, in most infill markets industrial rents still have room to double or triple or more before it becomes economically attractive to tear down other types of real estate and build industrial.

NIMBY forces are also strongly aligned against industrial real estate – ironically these forces are the strongest in markets where supply is already most out-of-balance with demand. Nobody wants to live near a distribution center with semi-trucks coming and going at all hours. For example, in the Inland Empire in Southern California, which is already the lowest vacancy and highest rent growth industrial market in the country with essentially zero available space, various municipalities either have or are considering moratoriums on new warehouse development.

Following the DRE acquisition, PLD’s portfolio is most heavily concentrated in these low-vacancy, NIMBY coastal markets: Southern California is their largest market at 21% of NOI, followed by NY/NJ at 9% of NOI and the SF Bay Area at 5% of NOI. Obviously if new supply is banned, rents must continue to grow rapidly – this is the rare real estate investment where nasty local politics are actually a tailwind for landlords (these same coastal markets all have some form of multifamily rent control and have flirted with commercial rent control).

The past two years of shrinking vacancy and skyrocketing rents have induced a historic supply response, with ~700mm SF of supply under construction. However, depending on which brokerage’s research you’re reading, somewhere between 30-50% of this has already been pre-leased. PLD estimates that if the entire un-leased forward pipeline delivered today fully vacant, national market vacancy would still be low enough to generate significant rent growth:

Additionally, 2022’s increase in interest rates and the subsequent drying up of construction financing has largely frozen new development starts. Once the current supply pipeline is absorbed, it is likely that there will be an ‘air pocket’ of minimal new supply beginning in 2024:

Finally, behind the headlines of Amazon downsizing its real estate footprint, the company is actually increasing the percentage of its footprint that it owns vs. leases. A company that truly believes its real estate footprint will be smaller over the long term should be doing the opposite: selling off its real estate and leasing it back so it can walk away from a much less valuable asset at lease expiration. 

Perhaps Amazon is looking at some form of the below charts, which show that industrial rent is still a relatively tiny percentage of overall distribution costs. With new warehouse supply constrained, there should be plenty of room within companies’ logistics margins for rents to multi-bag in the coming years.

Catalyst

Time / Compounding