12 September 2019

The yield compression cycle for Australian office markets that started in 2013 essentially came to a head around mid-2018. Since then, the Australia average prime CBD office yield has compressed just 10bps to June 2019. It could have been reasonable to assume that the yield compression cycle was over.

But then bond yields fell sharply and the base interest rate was cut 50bps to 1.0%, with the market and economists expecting further cuts, landing at a base rate of 0.5% by next March. With interest rates pushing to new lows, we forecast that the yield compression cycle has another leg to run and by end-2020 yields will be 25-50bps lower in most office markets. Re-pricing of office assets will also likely spill over to other property sectors.

Property Investors a Cautious Bunch


Direct commercial real estate is a high-priced, illiquid asset class; thus it’s understandable that investors take time to absorb new relative pricing paradigms before adjusting to them. Figure 1 partly demonstrates this point. The correlation between prime CBD office yields and bond yields is 0.61; however, if ‘noise’ is removed from the bond yield by trending it using a 24-month moving average, and then that series is lagged 12 months, the correlation rises to 0.82 over a twenty-year period. The key inference is that property pricing is slow to react to trend changes in bond pricing.

Figure 1: Prime CBD office yield versus 10-year government bonds

Office-Yield-Figure 1-no background
*Source: CBRE Research

Figure 1 also illustrates our expectations for bond yields to June 2022. We forecast that by mid-2020 the trended, lagged series will start to push materially downwards. If our bond forecasts prove correct and property yields haven’t compressed further by this stage, they will.

The case for yield compression

 

The simplest and most commonly argued case for yield compression is derived from comparing the spread between property yields and 10-year government bond yields. Figure 2 illustrates that the spread between the average for Australian CBD prime office has recently reached 460bps, well above the average spread of 380bps since 2011. We forecast that bond yields will average 1.4% to the end of 2022. Reversion to a 380bps spread over that 1.4% implies the current prime CBD office average yield of 5.7% could compress 50bps to 5.2%.  

Figure 2: Australian CBD prime office yield spread over 10-year Australian Government bonds

Office-Yield-Figure 2-no background

*Source: RBA, CBRE Research

Comparing bond total returns to property total returns tells a similar story. Using MSCI total return data from 1994 to 2018, we found that commercial real estate investors over the 24-year period received on average a 3.6% total return spread (aka property risk premium) over bonds if the property is held from 1 to 10 years (figure 3).

A startling revelation from our analysis was the consistency of the median total return spread over various hold periods from 1 to 10 years, with the median spread ranging from 3.46% to 3.75%. The distribution of the spread around median values reveals that holding property for longer time periods generates less volatile total return spreads.

Figure 3: All Property total return spread for various hold periods

Office-Yield-Figure 3-no background

*Source: CBRE Research, MSCI

How much yield compression can we expect?


An estimate of future yield compression can be made by assuming that markets will ultimately price commercial real estate so that reversion to the 3.6% total return premium occurs. For this exercise, a June 30, 2019 valuation of a B-grade Sydney CBD office building was used, with the following assumptions:

  • 5.25% cap rate
  • 5.75% terminal cap rate
  • 3.9% compounded annual growth of net rents
  • 6.4% Internal Rate of Return (proxy for total return)

We forecast that Australian Government 10-year bond yields will average 1.4% between June 2019 and December 2022. Adding to that 1.4% the 3.6% total return spread results in a target total return (IRR) of 5.0%. The DCF valuation was adjusted so that a 5% discount rate was adopted. We then adjusted the entry and terminal cap rates in the valuation. The results are presented in Figure 4, demonstrating that even if terminal cap rate softening of 100bps is adopted, a going-in cap rate of 4.6% will deliver the target total return of 5.0% - a 65bps compression from the current 5.25% on the asset modelled.

Figure 4: Cap rate compression scenarios

viewpoint_graphic_780px_300USETHISONE-01-01
*Source: CBRE Research

The 3.6% total return spread is an average derived from a sample that generated higher and lower spreads. It might be the case that the next few years will be a period where above-average spreads are recorded and yield compression is lower than in the scenarios modelled. For that reason we expect compression will range from 25 - 50bps.