Investigation soon proved, however, that the Harvard method was not readily applicable to New Zealand conditions. The Harvard index of business conditions measures “domestic” trade in the United States—there is available no suitable measure of “domestic” trade in New Zealand. The Harvard index measures monetary conditions by plotting changes in the rate of interest for short-term loans—in New Zealand there is little fluctuation in such interest or overdraft rates and changes in monetary conditions are registered rather in the variations of deposits, due to payments with the banks' London balances being credited to customers in New Zealand. New Zealand has no independent currency and banking system; but works under a Credit-Exchange standard which regulates monetary conditions here according to fluctuations in the London balances of the banks. And, as the present investigation shows, the relationship between these monetary conditions and speculative capacity is quite different also from that obtaining in the United States.

All these differences depend upon the fact that, in the United States, the main movements of business are normally self-contained and arise from domestic causes, while in New Zealand such movements are, in the main, transmitted from overseas.¹ Thus there is in the United States a genuine internal cycle of good and bad trade, but in New Zealand any cyclical fluctuations are merely the reproduction of similar movements in our overseas markets, transmitted to us through the credit-exchange standard which regulates our monetary conditions, and, with them, practically all economic activity in the Dominion.

In order to measure the movement of industrial share-prices, twenty-five companies were selected as representing different sections of the economic life of the community. In each case the ordinary shares of the companies concerned were taken as quoted in the local Stock Exchanges. The prices of these shares at the end of each month are tabulated from the reports of sales made upon the various Stock Exchanges. The price taken in each case is that at which most sales are made, or, in statistical language, “the mode.” Each share is then reduced to a percentage of its par value and the indexes thus obtained are combined in a simple unweighted index.

In the selection of items of the index various practical requirements had to be borne in mind. The essential thing was to get a fair variety and, at the same time, an adequate representation of the major industries in the Dominion. A rough weighting was carried out by the number of items of various kinds of shares included in the series. Regard was also had to the possibility of getting frequent quotations reflecting changing economic conditions.

After some experiment, twenty-five shares were selected, as follows:—4 Banks, 2 Insurance, 4 Loan and Agency, 2 Shipping, 2 Meat, 2 Coal, 1 Woollen Mill, 2 Building (Timber and Cement), 1 Sugar Company, 1 Drapery, 1 Drug Company, 1 Paper, 1 Co-operative Association, and 1 Gold-Mining.

The method of compiling the index may be illustrated by the following table:—

In gathering information over a fairly long series of years, certain difficulties were met with. Spells of infrequent sales were occasionally met in certain shares, necessitating recourse to interpolation. This was, however, not extensive enough to cause difficulty.

Index of 25 Industrial Share Prices 1919–1925, compared with Index of Banking Conditions (percentage of deposits to advances, etc.)

In the case of every share also there is a normal rise and fall before and after the payment of dividends, either yearly or half-yearly. Since the dividend payments are, however, spread fairly evenly over the months of the year, so that the rise in the value of one share “cum. div.” is offset by the fall in the value of other shares “ex div.,” any seasonal fluctuations arising from this cause tend to be smoothed out. In order still further to eliminate these regular seasonal variations, the final figures were plotted from a three-monthly average. The number of dividend payments falling within each quarter of the year is, March, 12; June, 13; September, 10; December, 15.

The chief remaining difficulty in measuring movements of share-prices due to changing business conditions arises from the periodic alterations of capital invested in different businesses. In certain instances calls have been made upon shareholders for further capital. This does not involve any great difficulty because the current price of

the share is then reckoned upon the new par value. Similar adjustment is made where share capital is written off as the result of accumulated losses. The opposite case presents a difficulty in calculation which the present index does not attempt to surmount, but which, for the purposes of the index, needs only to be mentioned in order to guard against a possible misinterpretation. Where a company capitalizes its accumulated profits by the distribution of bonus shares, there is a period of a few months during which investors discount the expectation of this distribution by offering higher prices for shares. For a shorter period just before the actual distribution there may emerge two prices for the shares—“ex rights” and “cum. rights”—while the rights to the new issue may be quoted separately. After the distribution of bonus shares has taken place, the share-value drops considerably and settles for some time on a much lower plane, which may even be below par in exceptional cases. The plotted indices of separate shares during the postwar boom period shows a series of increasing values rising to peaks and dropping suddenly.

The index which combines these movements of individual shares is, of course, smoother, since one movement tends to offset the other; but it follows, from what has gone before, that neither the individual level nor the combined index is to be viewed as a measurement of the relative values of shares over busy periods. It is not intended for that purpose; but has been devised purely to measure short-period movements of speculative activity. For this limited purpose its validity is not impaired by the occurrence of distributions of bonus shares, since these cause in themselves speculative activity quite legitimately included in any index devised for the purpose of measuring shortperiod changes in such activity.

If individual share prices are plotted separately there is, of course, a wide range of variation over a series of years. The many factors which influence different industries are complex and combine in different proportions to affect the value of the shares of individual company in very different ways. It is no part of the economist's task to investigate the conditions of particular businesses and still less to indicate the possibilities of getting rich quickly by speculative investment in particular industries. Any profits to be made in this way will be made after close study of individual businesses in their manifold aspects against a general economic background. It is hardly necessary to say that the proper person from whom to seek advice in the matter of investment is a competent sharebroker, not an economist.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

When, however, the vagaries of the movements of individual shares are combined in an index, there are certain interesting economic results which are perhaps best shown by comparison with the results of previous investigations, to supplement which the index was compiled. The following chart shows the index of share prices plotted immediately above the index of bank credit in New Zealand. The remarkable similarity of movement is evident at a glance and the coefficient of correlation between the two series worked out by Pearson's method r=σxy/n6₁ 6₂ is plus .95 with an infinitesimal probable

error of .013. This is proof of the coincident variation of these series, but does not, of course, establish any causal connection between them.

In other countries, and particularly in the United States, it has been shown that movements in the prices of industrial shares follow movements in bank credit and anticipate movements in domestic trade. No doubt if any reliable measure of domestic trade was available in New Zealand, it would be found to follow changes in share prices. Speculators anticipate forthcoming business developments, but the usual relation between bank credit and speculative activity is not present in New Zealand. The curves show rather that share-prices are slightly in advance of and certainly do not lag behind the banking index. This is the case, in the first fall of the slump period, in the middle of 1920, in the recovery of 1921–2, and in the slighter depression of the season 1923–4. Obviously movements in speculation are not governed by local banking conditions as in other countries, though they vary with these conditions. The facts suggest rather that both series are governed by an outside factor.

This lends further support, therefore, to the contention put forward at the beginning of this paper, viz., that changes in New Zealand economic conditions are transmitted from outside. It it natural to find that an appreciation of these changes occurs rather earlier on the Stock Exchange than anywhere else. The banking, instead of preceding first speculation and then general business conditions, moves along with speculation slightly behind it.

The peculiar facts of economic organization in New Zealand, would lead one to expect this. In the United States, the Stock Exchange borrows from the banks. The practice of daily settlement makes possible a much greater development of day to day borrowing at cheap interest rates than obtains even in London, where fortnightly settlement is still the rule and where day to day money is, therefore, absorbed rather by the discount market for bills of exchange. In New Zealand no such market exists at all. There is not even a day to day or week to week fluctuation in interest rates. The bank rate of interest on overdrafts, the only important interest rate in the Dominion from the point of view of short-term credit, has not changed now for nearly three years.¹

Instead of a free market for short-term credit in which a competitive demand is supplied to any extent at the current price, we have a regulated market in which price is fixed over long periods and the resources of the banks in New Zealand are, in effect, rationed.

In other words, the changes in bank credit are not determined by local conditions but local business conditions are determined by the bank credit available which in turn depends upon the London balances of the banks, which are built up as the net result of our overseas trade and overseas borrowings.

The relationships between different phases of economic activity disclosed by the Harvard investigations do not hold in New Zealand.

Instead of a fluctuating movement of local business affecting the bank credit available first for speculative activity and then again for general business, thus completing a fairly short cycle, we have to measure—

(a) The fluctuating relationship of imports and exports, both visible and invisible, the main determining factor in which is the tendency of export prices.

(b) The resultant London balances of the banks which operate in New Zealand.

(c) The transmission of these balances to New Zealand to alter the quantity of deposits in the local banks.

(d) The changed purchasing power thus made available which, in turn, affects imports and so adjusts the trade balance and the London bank balances.

This analysis, which shows the fluctuations of trade activity in New Zealand to be the consequential result of similar (and larger) fluctuations in great industrial countries, transmitted here through our credit-exchange system, is also a further piece of evidence against the crop-cycle theory which has been advanced to account for the trade cycle. In our case at least the trade fluctuations in a country which is predominantly engaged in the production of raw material and food-stuffs influenced by meterological conditions, are induced from industrial countries and through changes in demand affecting the prices of our exports. This supports the results of investigations which are at present being made at Cambridge by a former student of Canterbury College, Mr. H. Belshaw. He has come to the conclusion that the main determining cause of the economic rhythm or cycle which has attracted the attention of economists for many years and has been more widely discussed since the war, is not to be found in harvest variations connected with meteorological, cycles; but is due rather to fluctuations in industrial conditions which, owing to roundabout methods of production, take time to work themselves out, and therefore induce an irregular fluctuation of industrial production. Harvest variations may introduce complications and may even, at times, act as extra deciding factors; but our investigations of New Zealand conditions support Mr. Belshaw's contention that the main determining cause of the cycle must be sought elsewhere, and that it is changes in demand caused by industrial fluctuation in our overseas markets and causing changes in our export prices which is the proximate cause of cyclical fluctuations within the Dominion.