In recent weeks, financial markets have grappled with the reality that the Federal Reserve is moving to normalize monetary policy, and in the process, definitively bringing to an end the era of negative real interest rates. As a result, our clients are increasingly asking the obvious question: What will the impact of higher inflation and interest rates be on real estate pricing? Even for investors comfortable with today's prices and implied relative returns, a legitimate concern is the extent to which current attractive returns could be wiped out as the Fed normalizes monetary policy.
In this article, we take a look at the potential short-term and long-term impacts of higher interest rates on real estate total returns, cap rates and pricing. To be sure, there is no single "obvious" answer to the question about the impact of rising rates; the outcome depends on multiple interrelated assumptions. That said, if one accepts our J.P. Morgan Asset Management Long-term Capital Market Return Assumptions,1 which call for the 10-year Treasury to gradually settle at a 4.5% to 5.0% range over the next five years, we believe that today's private real estate pricing, which assumes unlevered total returns in the 6.5% to 7.5% range, is sustainable.
The short-term impact
It is important to separate the impact of higher interest rates into short-term and long-term effects. In the long term, the relevant question is: "Where do rates end up?" In the shorter term, the appropriate query is: "How will we get to higher rates?"
The road to higher rates can be steep, bumpy and without warning signs-as in the 1994 Federal Reserve tightening cycle-or less steep, orderly and well-marked, as was the path of 10-year Treasury yields in the 2004 cycle (Exhibit 1). These conditions can significantly influence the outcome for real estate capitalization (cap) rates in the short term. Clearly, the 1994 cycle resulted in a suboptimal outcome from the point of view of financial markets-one which the Fed is at pains to avoid this time around.2
The 1994 tightening caught markets off guard, leading to a spike in long-term Treasury rates. The impact was felt in the real estate market, but to a lesser degree than might be expected. The 10-year Treasury rose from 5.6% to 7.5% between 4Q1993 and 1Q1995, a 190 basis point (bps) increase. Yet transaction cap rates over that period rose by only 40 bps, from 9.0% to 9.4%.
In contrast, the 2004 tightening was well-telegraphed; Treasury yields actually rose ahead of Fed action and were well-behaved after the tightening began, rising a modest 4.6% to 5.1% over the 2Q2004 to 2Q2006 cycle. Transaction cap rates actually fell over that period, from 7.2% to 6.3%, as investors began to underwrite stronger income gains (and likely, lower risk premia) for real estate.
The "weak link" between changes in Treasury rates and changes in cap rates seen, to varying degrees, in these two tightening episodes, is not atypical.
We would expect the short-term response of cap rates to be similarly minimal in the coming tightening cycle, particularly given the expectation that the Federal Reserve telegraphs the rise in rates better than they did in the 1994 episode.
The long-term impact
Over the long run, the key question is: "Where do current real estate returns sit relative to some long-term equilibrium estimate of where those returns are likely to end up?" One way to address this question is through a broad capital markets framework. As Exhibit 3 illustrates, within this framework, core real estate, which has both bond-like and equity-like characteristics, should be priced to provide total returns somewhere between bonds and stocks. In addition, all risk assets should be priced to offer a return premium to "normalized"4 risk-free Treasury rates.
What is a reasonable assumption for a normalized Treasury rate? J.P. Morgan's Long-Term Capital Market Return Assumptions are a good starting point. Our strategists and market experts estimate that the 10-year Treasury yield will settle in at 4.5% to 5.0% over the next five years. The equity risk premium is assumed to be about 400 basis points. The core real estate premium, as one might expect, is assumed to be lower, at 200 to 300 basis points, consistent with history. Doing the rough math, "equilibrium" core direct real estate total returns should be somewhere between 6.5% and 8.0%.
To look at current pricing in this context, Exhibit 4 shows pro forma IRRs for core transactions underwritten over the past decade within J.P. Morgan's Global Real Assets (GRA) platform. The takeaway from this chart is that our underwriting IRRs have settled at around 7.0%. Is that 7.0% return sustainable or will it have to move higher when interest rates rise?
Exhibit 5 shows a simple analysis backing out the spread (in basis points) between current underwriting IRRs, in the top row, and "equilibrium" (or "normalized") 10-year Treasury rates, in the first column. In bold are the current range of IRRs being underwritten today (roughly 6.5% to 7.5% unlevered) and the range within which the our Long-Term Capital Market Return Assumptions project 10-year Treasuries to stabilize over the next five years (4.5% to 5.0%). This is essentially a "next-buyer" exercise, designed to answer the question: "If today's average core buyer sold in five years, at which point Treasury yields had risen to 4.5% to 5.0%, would the next buyer require a higher return to be consistent with long-term return premia?"
The answer-even at a 7.0% IRR-would generally be "no" given a 5.0% Treasury yield, as that would still provide a 200 basis point spread (though pricing would be tight, as this is the low end of historical required spreads). However, if Treasury rates end much higher than 5.0%, required returns would have to move higher to achieve what has historically been viewed as a sufficient return premium. If expected income growth at that time does not rise commensurately, cap rates will have to rise and prices fall. So part of determining whether one is comfortable that today's return level is fair and sustainable is staking a claim on where rates end up; a 5.0% Treasury world is tight, but a 4.0% world would leave real estate still looking very attractive on a relative return basis.
There is some built-in conservatism in underwriting
In assessing the degree of "cushion" against rising rates given current real estate values, it is important to understand the assumptions behind today's IRR calculations. For example, in anticipation of rising rates, today's IRRs, in general, already incorporate an assumption of higher future cap rates. Exhibit 6 illustrates this point by clarifying the exit cap rate assumptions behind discount rates (appraisal IRRs) on real estate held in three GRA open-end funds covering the spectrum from core to value-added. Discount rates range from roughly 7.4% to 8.0% across strategies. In each case, exit cap rates are assumed to be higher than initial yields by anywhere from 65 to 180 basis points. If those higher exit cap rates were not assumed, unlevered total returns would be close to 8.8% and 8.3% for core and core plus respectively and close to 10.4% for value add.5 The cushion in today's values, therefore, may be even wider than IRRs that incorporate the assumption of higher exit cap rates suggest.
Competing spreads are not yet signaling "danger" for real estate IRRs
A further necessary condition for a given level of underwriting returns to be sustainable is a reasonable spread to competing assets and to mortgage rates. Exhibit 7 shows the spread of core underwritten IRRs to BBB bonds and to our proprietary index of 65% loan-to-value (LTV) commercial mortgage rates. In the last tightening cycle, these spreads approached historically narrow levels as early as 2006, portending a rise in core IRRs. Today, both spreads, which speak directly to the ability to obtain a mortgage and the relative return of similar risk assets, are nearly as high as they have ever been.
What's the takeaway?
Putting together two concepts: (1) unlevered returns relative to current competing returns and (2) unlevered returns relative to "equilibrium" Treasury rates, we can devise a way to judge whether today's returns are sustainable or not. A reasonable rule of thumb is that as long as there is a decent core spread to bonds and debt, current return levels are very supportable with a 10-year Treasury that settles in the low 4% range, tight but OK if we end up closer to 5%, and pressured to move higher (and prices lower) at a level much above 5%.
Caution is still called for
The forgoing thoughts are not an argument for investor complacency, however. A 10-year Treasury that ended up at 4.5%-5.0% still makes a lot of 6.0% IRR underwritings-which are occurring in the market-look very tight, one might say priced to perfection. Even at 7.0% IRRs, there are headwinds. In particular, the cost of debt would probably end up near-or a bit above-today's 5.0% to 5.5% income yields, making debt only marginally accretive in many cases. Even when debt remains accretive, the easy levered core strategy that today routinely produces 9.0% to 10.0% levered returns would fade away as levered returns converge toward unlevered returns. In this context, it makes sense to think of today's anomalously low interest rate environment as a (welcome) windfall to current levered returns that will eventually moderate, exposing both underperforming and overpriced assets along the way.
- J.P. Morgan Asset Management Long-term Capital Market Return Assumptions-2013 estimates, September 30, 2012.
- Will Fed tapering lead to a repeat of the 1994 sell-off?, J.P. Morgan, May 15, 2013.
- Cap rates are one-year ahead income yields: cap ratet=NOIt+1/Pt. If one makes the simplifying assumption of constant income growth (a common industry shorthand to a full DDR model), the cap ratet equals required returns (RRt) minus expected property income growth (E(gNOI)): cap ratet=RRt-E(gNOI). So even if required returns went up in complete lockstep with interest rates (something we argue further on should not be expected), an equal rise in income growth expectations would leave cap rates unchanged.
- Normalization refers to the restoration of economic conditions, such as interest rates, to more cycle-neutral levels, following a temporary dislocation period.
- Those IRRs typically assume 10-year holds and are built up as initial yield + annual cash flow growth-impact of higher exit cap rates. The higher IRRs are calculated by assuming exit yields=initial yields.