Interest and Non - Interest Terms in the Process of Mortgage Market Clearing

نویسندگان

  • Ciaran Cassidy
  • Gabriel Fagan
چکیده

The paper is devoted to a test of the hypothesis of mortgage market disequilibrium in Ireland. The mortgage rate is found to be very sluggish upwards when excess demand prevails, lending support to the mortgage rationing theory. In an incipient excess supply regime the mortgage rate is lowered to the market-clearing level instantaneously. We also obtain evidence that the down-payment ratio is used by building societies as a sorting device. The role of this non-price term creates econometric difficulties in testing for disequilibrium. These difficulties are dealt with in a novel way. I I N T R O D U C T I O N A debated topic in the economic literature relating to the housing and mortgage markets concerns the precise nature of the interlinkages between these two markets. The essence of a long-standing conventional wisdom is that mortgage rates, by themselves, do not fully reflect the avail­ ability of mortgage credit. The mortgage rate is sluggish to move in response to perturbations in supply and demand and therefore rationing and, perhaps occasionally, excess supply may exist and impinge directly on the housing market. However, despite the ubiquity of this belief there are few rigorous econometric tests of the disequilibrium mortgage market hypothesis. It would appear that the conventional wisdom has come to be accepted as virtually self-evident due merely to its repeated affirmation. The amplitude of the trade cycle in the residential construction sector is much greater than that for the economy as a whole. Indeed, it could be *The paper has been considerably improved in the light of two excellent referees' reports and the comments of Michael Moore, John Frain, Tom O'Connell, Dan Seidmann, Ciaran Cassidy, Gabriel Fagan and Patrick Carey. argued that the trade cycle in homebuilding is probably one of the major contributors to the overall trade cycle. This is a world-wide phenomenon and not exclusive to Ireland. It has prompted policy-makers to search for counter­ cyclical stabilisation policies. In the U S , for example, the academic diagnosis of the problem continues to be fairly well represented by the following quote from Friend writing in 1969: "The greater impact of monetary stringency on housing than on the rest of the economy apparently is due to a capital rationing effect, resulting from deficiencies in current institutional arrange­ ments for providing mortgage credit . . ." (1969, p. 8). An indisputable fact in the Irish case is that the mortgage rate of interest is sluggish. Over the time span of the present study, 1968 to 1983 inclusive, the mortgage rate changed only 26 times in a total of 64 quarterly observa­ tions. The rate remained unchanged for almost four years from 1969 Q2 to 1973 Q l . Structural and cost factors undoubtedly play a role in explaining this sluggishness. Another contributory factor is the fact that the mortgage rate is a very politically sensitive variable. Of course, the mortgage market is always cleared ex post in the sense that there is a unique quantity of mortgages traded. Since the mortgage rate does not vary sufficiently to ration potential customers within the quarterly observation unit of the study, nonprice rationing must be used to allocate the available flow supply of mortgages. It is hypothesised that the dominant non-price variable working to this effect is the downpayment ratio. Non-price rationing creates special econometric difficulties in estimating the extent of price rationing. We deal with these below. Measures of proxies of mortgage availability used in empirical models of the housing market tend to be somewhat ad hoc and to lack rigorous justifi­ cation. A related drawback of these proxies is their failure to distinguish between temporary and permanent effects of mortgage market spillovers on the housing market. Does "availability", for example, have an enduring effect on housing demand and thus on house prices or does the effect only endure for the amount of time required for the mortgage market to clear? Yet another related deficiency of conventionally-used proxy measures, whether it be changes in the stock of saving accounts available to Saving and Loan Associations (see, for example, Jaffee and Rosen (1979)) or changes in the stock of mortgages (see, for example, Hendry (1984)), is their lack of symmetry. These proxy measures presume that the mortgage market is characterised by perpetual excess demand. If the market is in equilibrium or in excess supply, then these proxy measures have no constraining effect on housing expenditure and hence simply introduce unnecessary noise into the regression analysis. Available evidence would suggest that monetary policy in Ireland has no enduring effect on those areas of activity which are financed by bank credit. This is because money and credit are internationally tradable assets and hence an artificial shortage (surplus) of these created by domestic monetary policy in the interest of pursuing some domestic macroeconomic objective will be negated by the international movement of liquid claims (see Browne, 1986). The one area of economic endeavour which may have to be excepted from this conclusion relates to the mortgage and housing markets. Mortgages are generally not tradable internationally. Hence mortgage market disequilibrium cannot be offset by a supply of close substitutes from abroad. Therefore one of the major channels of transmission of monetary disturbances, whether of domestic or foreign origin, is via mortgage market disequilibrium to the housing market. The fact that the housing market is subject to such huge cyclical variation in activity may be due in no small way to the fact that it is a hostage to those monetary shocks. The main purpose of the present exercise is, therefore, to test for price clearing in the mortgage market. Recent progress in the theory of market disequilibrium and in the econometrics of disequi­ librium estimation have now rendered this task feasible. The structure of the paper is as follows. In Section II we present our approach to estimation of markets in dynamic disequilibrium. This stems largely from the work of Bowden (1978). Section I I I contains our fleshed-out specifications of the mortgage supply and demand schedules. The fact that mortgage demand is predominantly a derived demand from the demand for private sector housing is fully acknowledged by the inclusion of the arguments affecting housing demand in the mortgage demand schedule. Section I V is devoted to a presentation of our results and also deals with the phenomenon of non-price rationing and how this affects our estimation of price rationing. This is followed by an evaluation of our results and comments on the speed of adjustment of the mortgage rate to its market-clearing level and also on our supply and demand elasticity estimates. Section V contains some conclusions. II D I S E Q U I L I B R I U M E C O N O M E T R I C M O D E L A standard Fair-Jaffee (1972) characterisation of the non-market-clearing model of the mortgage market runs as follows: M? = X j o , + a , r m t + e 1 ( ; (1) S a i ^ a 2 r m t + e u > M t = m i n ( M ? , M S ) , ( 3 ) A r m t = X ( M ° M ^ X 1 > 0, (4) X and Z are vectors of exogenous variables affecting demand and supply respectively, r m is the mortgage rate of interest and e t and e 2 are stochastic error terms which are individually normally distributed and serially indepen­ dent. They are also independent of each other. Equation (3) is the short-side rule which says that the quantity transacted is the minimum of supply and demand. The model is completed by Equation (4) which provides a regime indicator with an increase (decrease) in price indicating excess demand (excess supply). In Bowden's (1978) formulation he replaces Equation (4) with the follow­ ing specification: r m t = " r m t l + ( 1 " ) r m t > ( 5 ) i.e., the current actual observed interest rate (r t ) is a weighted average of last period's rate ( r ^ . ^ ) and the current market-clearing rate ( r m * ) . A deficiency of the formulation in (4) is that there is no well-defined test for market clearing which occurs when X = °°. In other words, the market-clearing hypothesis is non-nested. A test of the market-clearing condition within the Bowden formulation is straightforward (i.e., a test of n = 0). ju = 1 implies infinitely slow clearing. Charemza (1979) and Ito and Ueda (1981) have improved upon the Bowden formulation by distinguishing between upward and downward price flexibility as follows : r m t = ^ l r m t l + ( 1 ^ l ) r m t * ' i f r m t > r m t l > ( 6 a ) r m t = ' I 2 r m t l + ' i 2 ) r m t * ' i f r m t < r m t l ' ( 6 b ) This is the procedure we employ here to test for market disequilibrium. Thus our model consists of Equations (1), (2), (3), (6a) and (6b). Invoking the short-side rule (Equation (3)), we can integrate Equations (6a) and (6b) into the structural equations for demand and supply (Equations (1) and (2)) to yield the following two equations (see Ito and Ueda, 1981, p. 695): ^ = X i a i + a 2 r m t + « s A C + e i « . ( 7 ) M . a z ! f l i + v « , + M r ; « + e ! , . w 1. A more general formulation of Equations (6a) and (6b), as well as Equations (5) and (4), would be to set these equations in a stochastic setting by adding error terms. Furthermore, a still more general model could be had by allowing for inter-temporal spillovers from past disequilibria in the market. M O R T G A G E M A R K E T 75 where the regime switch variables ( A r m t and A r j n t ) are defined as follows: A r + = J r m t " r m t l > i f r m t > r m t l and where: m t f zero , otherwise Ar r m t r m t l ' l f r m t < r m t " l mt zero , otherwise Estimates of upward and downward adjustment speeds (j&j and £ 2 ) c a n be derived by substituting for the estimated values of the parameters in the Equations in (9) and estimates of their respective standard errors can also be obtained by using the formulae given in Ito and Ueda (1981, p. 696). It should be noted that Equations (7) and (8) relate to the entire sample period and not just to periods of excess demand and supply, respectively. Equations (1) and (2) above refer to the long-run optimal amounts demanded and supplied. These can clearly differ from the short-run optimal amounts when costs are incurred in adjusting asset quantities. Browne (1987) has argued that a defect of the above disequilibrium estimation model is that it makes no provision for sluggish quantity adjustment. This omission can result in biased estimates of excess demand and supply and, in some cases, can lead to the mistaken identification of an excess demand regime as one of excess supply and vice versa. The probability of incurring these biases can be greatly reduced by allowing the model to accommodate sluggish quantity adjustment. We do so here by postulating independent generalised stock adjustment mechanisms governing both the demand and supply of mortgages. (For an application of this approach with Irish data see Browne and Honohan, 1988.) A detailed account of this approach is postponed to the following section. I l l D E T A I L S O F M O D E L S P E C I F I C A T I O N The model's structural equations are given in Equations (10) to (15) in Table 1. Equation (10) is long run or desired stock supply of building societies' mortgages. It is an increasing function (a t > 0) of the difference between own yield ( r m ) and the opportunity cost yield of mortgages. The latter ( r I B ) is assumed to the three months interbank rate of interest on the Dublin money market. The term V(Ph) is a measure of the variability of house prices. For a given size of mortgage the higher is V(Ph) the greater is the probability that the market value of the average house will fall below that of the average mortgage and hence the riskier becomes the building societies' investment. For risk-averse building societies, a 2 < 0. Mortgage supply is also an increasing function (a 3 > 0) of repayments on existing mortgages ( R E P ) . The data for R E P are calculated as the difference between gross mortgage advances and the change in the mortgage stock in the current period. Total resources available to building societies is denoted by S A R . It is the sum of share and deposit accounts. Finally, Ph is an index of new house prices. Table 1: Model Structure (M/Ph)f* = a 0 ( S A R / P h ) t + a i ( r m r I B ) t + a 2 V ( P h ) t + a 3 ( R E P / P h ) t + u h (10) (MA/Ph) t = 8 1 [ ( M / P h ) S * ( M / P h ) ^ ] + 6 2 ( L R L R ) t _ j + 8 3 D R t + u 2 t (11)

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Interest and Non-Interest Terms in the Process of Mortgage Market Clearing: A Reply

horn (1990) argues that, in a situation of excess mortgage demand, loan JL applicants "with relatively low LV ratios (high DR ratios) have their demands satisfied while those with relatively high LV ratios are placed in a 'mortgage queue'. Hence successful applicants are on their desired demand curves but the aggregate demand curve does not shift and excess demand remains a feature of the marke...

متن کامل

The Relationship Between the Facility Interest Rate and Three Main Variable of the Money Market In Iran (1986-2017)

The bank interest rate is one of the most important macroeconomic variable in each country economic. The purpose of this paper is find the relationship between the facility interest rate and three main variable of the money market.In Iran. this issue for equations the interest rate facility, the interest rate of deposit, inflation and credit risk utilizing the model simultaneous equation system...

متن کامل

تزاحم منافع در داروخانه

The conflict of interest is a situation in which professional judgment and performance in the primary interest and obligation tend to be unduly influenced by a secondary interest. Pharmacy is one of the main rings in the process of providing healthcare services. In this process conflict of interest may occur frequentlywhich occasionally has influence on the professional and moral duties of phar...

متن کامل

Portfolio Optimization and Mortgage Choice

This paper studies the optimal mortgage choice of an investor in a simple bond market with a stochastic interest rate and access to term life insurance. The study is based on advances in stochastic control theory, which provides analytical solutions to portfolio problems with a stochastic interest rate. We derive the optimal portfolio of a mortgagor in a simple framework and formulate stylized ...

متن کامل

The Effects of Interest Rates Volatility on Stock Returns: Evidence from Bangladesh

The paper investigates the effects of interest rates on stock market performance by using monthly time series data for the economy of Bangladesh over the period of 1991 to 2012. A wide range of econometric techniques have been employed to analyze the relationship between the interest rate and stock market return. The study reveals a stable and significant long run relationship between the varia...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2012