Long-run equilibrium for the Greater Paris office market and short-run adjustments

Abstract

While there is an abundant literature on office market modelling, only few studies focus on the Greater Paris case, which stands as the largest office market in Europe. This article aims at contributing to the understanding of the Parisian rental office market underlying mechanisms. We use cointegration techniques including Gregory–Hansen structural break approach to model the market over the period 1990–2013. We employ a two-stage error correction framework to identify the long-run equilibrium rent as well as the short-run adjustments in rent, vacancy rate and stock. The findings show that once allowances are made for a regime shift in the long-run rent relationship, the specified model explains relatively well the Parisian market. The long-run results confirm the role of employment and total stock as long-run determinants of rental values. Yet, in the short-run, due to the inertia arising from the intrinsic features of the office market, lagged dependent variables appear to be drifting rents away from their equilibrium.

Keywords:

• Greater Paris office market
• error correction model
• Gregory–Hansen
• structural break
• French commercial property

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. This abundant literature includes studies such as De Francesco, 2008; Englund, Gunnelin, Hendershott, & Söderberg, 2008; Hendershott, 1995; Hendershott, Jennen, & MacGregor, 2013; Hendershott, MacGregor, & Tse, 2002; Ke & White, 2009; McCartney, 2012; Rosen, 1984; Wheaton, 1987; Wheaton, Torto, & Evans, 1997, etc.

2. This is indeed the case as we have $2\%\le {\mathit{\nu }}_t\le 11\%.$

3. We have not adopted a multivariate approach in the lines of Johansen (1988, 1991) because our cointegration relationship includes a structural break. We rather adopt a two-step approach as in Engle and Granger (1987) and estimate each of the short-run equations after estimating the cointegration relationship as explained in detail below.

4. When estimating the parameters of the cointegration relationship, problems arise concerning the asymptotic distribution of the OLS estimators. This is due to the presence of long-run correlation between the cointegrating equation errors and regressor innovations, on the one hand, and to cross-correlation between the cointegrating equation errors and the regressors, on the other, when the regressors are not exogenous. The FMOLS approach of Phillips deals with this issue.

5. If the series are cointegrated, ordinary least squares (OLS) estimation of the cointegrating vector is consistent, converging at a faster rate than in the standard case (Hamilton, 1994) and Phillips and Hansen (1990) employ a semi-parametric correction to solve the previous problems.

6. Two additional remarks can be made regarding the comparison between the Greater Paris and London markets: First, Hendershott et al. (2002) find a higher estimate of price elasticity for the London market, at −0.54, questioning the lower elasticity compared to the Paris market. This result confirms our previous concern that caution is required when comparing different studies. Because it is more recent, we use Hendershott et al. (2010) as a benchmark. Second, the geographical scope of the two markets is also likely to impact the elasticity calculations. In particular, the Greater Paris market displays larger boundaries compared to the London market.