by Neil Favanger
With the construction industry reeling under its worst recession for 10 years, if not its worst ever, the key question which is being asked by all sectors of the industry is how much longer will it continue?
There has been much speculation on this, but an investigation and analysis of currently available market data and statistics, combined with a measure of interpretation, can provide some much-needed indications of where the industry is now and where it is heading.
This article examines the long-term trend in tender-price inflation movements for building works and the cyclical nature of building tender inflation. By tender-price inflation movements, we are talking about the rate of increase (or decrease) in tender-price levels, rather than the movements in absolute tender-price levels themselves. While, in the long term, tender-price levels have steadily increased, the rate of increase (the inflation) has varied.
It is important to remember that if tender-price inflation is decreasing then absolute tender-price levels can still be on the increase! So long as the level of tender-price inflation is greater than zero then tender prices are rising. On a graph showing tender-level inflation:
- If the line is rising and is above zero, the rate of increase is accelerating;
- If the line is falling towards zero, the rate of increase is decelerating;
- If the line is falling below zero, the rate of decrease is accelerating;
- If the line is rising up towards zero, the rate of decrease is decelerating.
This article seeks to demonstrate a “cycle” in building tender-price inflation and from it to make predictions regarding the length of the current construction recession and the time remaining until the upturn in construction activity arrives. If a cycle can be shown to exist — or perhaps more fairly, if the statistics show that a cycle correlates fairly well with the reality, then this prompts the thought that if one can predict when the construction recession will end, when should one aim to build?
The factors which affect the state of the construction market in general, and building tender prices in particular, are many and varied. The interrelations between the factors are even more complex, and it would be impossible to take one single contributing factor in isolation and try to make predictions based on that factor alone.
However, it is fairly easy to demonstrate graphically how some of the major components which affect building tender prices interlink. The changes in contractors’ output, employment in construction and general building costs are probably some of the more significant factors affecting building tender prices, and these three also incorporate many of the effects of other factors as well — interest rates, for example. Using figures for contractors’ output rather than new orders has the advantages of making some allowance for the amount of work that contractors actually have on their books.
Figure 1 shows the average movements in output and employment in the construction industry with movements in general building costs and building tender prices over the past 15 years. It also includes forecasts for 1991 onwards where these are available.
The data used in the graphs has been prepared by, first, calculating year-on-year percentage increases in the factors and then calculating four-quarterly moving averages to smooth out the effects of any seasonal variations in the figures.
The graphs begin in 1975 with rising inflation in both output and employment, cost increases having just peaked, and tender levels showing signs of a rise. By the end of the 1970s, output, employment and tenders had passed their peaks and building costs were on the increase.
The slump of the early 1980s is an obvious feature of the graphs, with output and employment taking a nosedive, followed by building tender prices. After a slow recovery tenders levelled out and it was only as output and employment rose during 1986 and 1987 that the pressure on tender levels again caused a rise.
In 1990 the market crashed again; but the signs had been visible in the statistics for some time previously. Output had peaked in 1988, employment had just reached a high and building costs had been showing a gentle rise for over a year.
There was a limit to how long the industry could last on its “work in progress,” although when it came the crash was far greater than could have been expected. Or was it?
Movements in inflation
It has often been said that the construction industry moves in cycles, and there have been many investigations into the factors which affect the construction market and whether there are any cycles that these inputs exhibit. However, there seems to have been little analysis of the frequency of these cycles where building tender prices are concerned.
In order to examine the movements in building tender-price inflation it is necessary to use as long a time period as possible to evaluate the data. Figure 2 shows the annual percentage change in building tender prices for each year since the mid-1950s.
The basic input data for the graph is a synthesis of data from BCIS and other sources. As before, calculations have been made to produce annual percentage increases.
On the same axis as the graph of building tender-price inflation an oscillating “activity cycle” has been plotted. The activity cycle has been generated to best match the changes in the inflation in building tender-price levels over the period shown, but the cycle keeps its frequency and amplitude constant.
The cycle has been generated with a wave frequency of just under eight and a half years, that is to say, there is just under eight and a half years between successive booms (and consequently between slumps). A wave peak-to-trough amplitude of some 12% fits the tender peaks and troughs fairly well, and also produces the best fit statistically. While the two graphs do not mesh together precisely, the inherent peaks and troughs in the tender inflation graph match the activity cycle fairly well.
It would be folly to suggest that the maximum “swing” in tender-price inflation will never exceed this 12% figure (the graph shows that it already has on several occasions) — it is perhaps better thought of as a band within which tenders could be expected to remain most of the time; a sort of broad-band exchange rate mechanism.
The important thing to consider is how the peaks and troughs in actual building tender-price inflation fit the corresponding peaks and troughs of the cyclical graph.
Figure 3 shows the same graph of building tender-price changes as before, but this time overlaid with a line representing the same price changes, adjusted to remove the eight-and-a-half-year cyclical effect. The overlaid line shows the underlying trend movement in much the same way as seasonally adjusted figures smooth out peaks and troughs from sample data.
The trend line, while not being flat, is certainly well smoothed from the original data, and shows an underlying trend that rises from the late 1950s to a peak in the latter half of the 1970s, from whence it has been dropping back slowly. The fact that the overlaid trend line with the cyclical effect removed has flattened the data so well demonstrates that the activity cycle does indeed closely fit the facts — there are no large divergences from the general movement in the underlying trend.
The trend line also shows that since the mid-1950s there has always been an underlying rise in tender inflation. The bottom “limit” of the trend line is around the 4% to 5% mark, though it clearly rose much higher in the late 1970s, during the period of the highest levels in general inflation in the economy. It is only recently that there have been times when tender levels have fallen. Despite these, the underlying trend has been an increase. So, if there is a “set” cyclical element to building tender-price inflation movements, it should enable us to predict the short-term future of the market.
The first thing that Figure 2 indicates is that the recession in construction activity, and therefore in tender levels, is far from over, and this is supported by the available data. One has only to look at the various “state of trade” inquiries from the contracting organisations and the material producers to know that we have yet to turn the corner. The latest available employment statistics, and the forecasts of contractors’ output also give little hope of an early recovery.
The cycle indicates that there will be at least another year of recession in construction activity before things show a noticeable improvement, and building tender-price inflation is unlikely to show a real rise until 1993.
Short-term future
So, having completed our analysis of the cyclical pattern underlying tender-price inflation, what conclusions can we draw in terms of the short-term future of the construction market?
It would appear that the market is going to stay “in the dip” for at least the next year, while the recession in construction bottoms out. Thus we can expect a period of reasonably stable tender-price levels.
Certainly, the commercial property sector may well take some considerable time to recover from the current oversupply, especially in London. Other sectors may be far quicker to recover — the housing sector is usually more responsive to changes in the general economy.
The key question in the commercial market when looking for the ideal time to procure construction work must be how to achieve that combination of being able to build cheaply but let or sell into a rising market.
There are identifiable dangers of building both at the start of a recession and when the market is coming out of one. The difficult period at the start of a recession has now been passed: the time when contractors are having to reduce tender levels to maintain turnover against a background of increasing costs — providing the scenario which gives the greatest likelihood of “claims contracting”. Once the recession ends and tender opportunities start to return, the resulting rise in tender prices will cause contractors to give a lower priority to those “break-even” projects won during the depths of the recession, compared with those now available where the potential returns are so much greater.
In light of this, the ideal time to build, if there really ever is such a thing, is when market risk and variation is at a minimum — when the market is relatively stable during the period at the bottom of the construction activity cycle.
We are entering this “bottom-out” period of the present building tender-price inflation cycle and, as such, now is probably the ideal time to procure building contracts which will achieve optimum value for money while largely avoiding the disruptive factors of a fluctuating market.
The ideal for a successful development is to buy land cheaply, fund competitively, build cheaply and let or sell into a rising market. The first three components of this equation are now largely in place — although the availability of property funding will perhaps remain difficult for some time yet. The real trick, of course, is also to identify the timing of the upturn in demand for developments. For those developers who are prepared to risk a prediction that the upturn will come in the next two years or so, then now is the time to consider procuring building work.
In summary, with the “activity cycle” indicating at least another year of recession before things begin to show a noticeable improvement, the conclusion would have to be to build now while the recession lasts, provided you can get the funds and believe that you will be able to dispose of the development on completion!
