Key Trends in Globalisation

The statistical data indicates the population and shareholders have counterposed interests - implications of the negative correlation of economic growth and share prices

Summary

The studies on the relation between per capita GDP growth and returns on shares persistently find that the two are negatively correlated - i.e. a higher rate of growth of per capita GDP is associated with a lower return on shares. Cardiff Garcia and Neil Hume, in the Financial Times, recently logically used this data to deal with the relatively narrow issue of questioning analyses that the high growth rates of emerging markets, for example China, will lead to them displaying superior share performance to developed markets. But the implications of the overall finding are much more widespread and fundamental for economic policy. Specifically the implications of a negative correlation of per capita GDP growth and returns on shares include:

• The population's living standard, i.e its consumption, is on average positively determined by, and positively correlated with, growth of per capita GDP. However the return on shares is negatively correlated with per capita GDP growth. It follows that the interests of the population's living standards and the interests of shareholders are counterposed - the population has an interest in higher per capita GDP growth whereas shareholders, in terms of return on shares, would benefit from lower per capita GDP growth.
• Governments targeting high rates of increase in per capita GDP, that is high increases in living standards, should expect lower returns on shares than those with low rates of increase of per capita GDP, that is living standards. Therefore, high rates of per capita GDP and living standards, and high rates of growth of share prices/total return on shares, are counterposed and not complimentary objectives.
• These issues are in addition to the points made in the Financial Times that over the long run the higher economic growth rates of emerging economies would produce lower growth of share prices and total return on shares - this clearly fits China which has the highest rate of growth of any major economy but an underperforming share market. 

It should be stressed that such processes operate 'other things being equal' and over long periods of time. But they provide a long term determinant of economic developments which underlays, and therefore inflects, more short term trends. 

Given the widespread implications of this issue the present article therefore outlines the data and examines in more detail some of the implications.

*   *   *

Not just emerging but developed economies

The Beyond BRICs section of the Financial Times recently analysed a report by Deutsche Bank's John-Paul Smith which argued that share prices in emerging markets will suffer negative trends due to these countries government policies. The article however failed to provide the essential context that a negative correlation exists between per capita GDP growth and share price growth and total return on shares not only in emerging but in developed economies. It therefore follows, other things being equal, that high per capita GDP growth in some emerging markets would be anticipated to be correlated with underperformance of their shares.

However, this clearly raises a more general issue. Given that the aim of economic policy is to increase sustainable GDP per capita, as this makes possible increased living standards, this goal necessarily means that governments in both emerging and developed economies have to target high GDP per capita growth even although this will be necessarily be associated with lower share price growth and lower return on shares. The opposite course, targeting higher share prices and higher return on shares at the expense of lower GDP per capita growth, and therefore lower living standards, would mean placing the interests of the population lower than that of shareholders - a goal which, if being pursued, should be explicitly stated and would be unlikely to command widespread support. 

More specifically John-Paul Smith criticised governments of emerging markets on the grounds that: ‘GEM (Global Emerging Markets) companies are increasingly seen as “facilitators” for the broader economy, on the Chinese model, across many GEM markets.’ Stefan Wagstyl of the Financial Times paraphrased Smith as arguing: ‘The state will drive the redistribution of resources away from shareholders and in favour of labour.’

However it is well established in the studies on the correlation between share price growth/total return on shares and per capita GDP growth - for example Siegel's Stocks for the Long Run, Dimson, Marsh and Staunton’s Triumph of the Optimists and Ritter’s Economic Growth and Equity Returns - that a negative correlation between per capita GDP growth and share prices exists in both developed and developing economies. Such negative correlation is not due specifically to government policies in emerging markets but is an overall feature of economic growth.¹ 

The data

Turning first to the data establishing the negative correlation between GDP per capita growth and share prices, Siegel found: 'Real GDP growth is negatively correlated with stock market returns. That is, higher economic growth in individual [developed] countries is associated with lower returns to equity investors. Similarly, the stock returns for the developing countries against their GDP growth are plotted in Figure 81-b Again... there is a negative relation between the returns in individual countries and the growth rates of their GDP.' (Siegel p124) Siegel's findings for developed economies are shown in Figure 1 and those for developing economies in Figure 2.

Figure 1

Figure 2

 

Dimson, Marsh and Staunton found, regarding the total real return on equities and GDP per capita growth: ‘statistically, the correlation is -0.27 for 1900-2000 and -0.03 for 1951-2000.' (Dimson et al p156)

Regarding the correlation of GDP per capita growth and share price increases Ritter found: ‘My calculations for… 16 [developed] countries over the 1900-2002 period get a correlation of -0.37.’

In a recent survey Jain and Kranson's ‘The Myth of GDP and Stock Market Returns’ noted: ‘The data shows clearly that, over long periods and when adjusted for inflation, stock market returns and GDP per capital growth are negatively correlated.’

Goldman Sachs private wealth group, in a review, recently noted that the overall negative correlation between GDP per capita growth and share price growth extended to different groups of economies when ranked by growth rate: ‘Dimson et al have shown that if one invested in the slowest growing quintile of countries during this hundred-year- plus period, the equity returns would have outperformed the fastest growing quintile by 3% a year… Our own analysis for emerging market countries since 1991 showed the equity markets of the slowest growing countries within emerging markets outperformed those of the fastest growing countries by nearly 5% a year.’

This relation found by Goldman Sachs is shown in Figure 3.

Figure 3

 

China

Goldman Sachs noted China as illustrating this general trend: ‘China probably provides one of the best examples of the lack of correlation between strong economic growth and equity returns… China’s economy has outgrown that of the US by about 8% a year since the end of 1992 (the inception date of the MSCI China equity market index). Its equity market, however, has lagged that of the US by about 8% a year. Over the last 15 years, earnings per share growth in China has been negative 0.9% while that of the S&P 500 companies has been 5.4% a year. Most recently, in 2010, China has outgrown the US by an estimated 7% but the MSCI China Index has returned just 4.8% (the local Shanghai Composite Index is actually down 12.8%). On the other hand, US equities have returned 15.1%. Since the peak of US and Chinese equities in October 2007, China has outgrown the US by an estimated 10% a year, but Chinese equities have lagged the US by 2.7% a year.’ This data is shown in Figure 4.

Figure 4

 

Goldman Sachs survey therefore concluded: ‘Whether it is 1 year, 3 years or 18 years, economic growth has not translated into better investment returns in China … The evidence shows that faster economic growth rates do not result in higher equity returns. In fact, if faster growth is priced into the equity markets, the equity markets are most likely going to lag those of slower growth economies.‘

In a recent update of their study, for the Credit Suisse Global Investment Returns Yearbook, Dimson, Marsh and Staunton concluded: ‘[Our] Figure 2 ranks the real equity return of the 19 Yearbook countries over the period 1900–2009, from lowest to highest…. [our] Figure 2 shows that the supposed association between long-run real growth in GDP per capita and real equity returns is simply not there (the correlation is –0.23).’ This is shown in Figure 5.

Figure 5

 

Dividends, share prices and growth

In their study Dimson, Marsh and Staunton however noted an extremely close correlation between share prices/return on shares and dividend payments: ‘there is a high correlation (0.87) between real equity returns and real dividend growth across the 19 countries.’ 

Such a close correlation between dividend payments and share prices is related to the negative correlation between share prices and economic growth: ‘the claim that real dividends grow at the same rate as real GDP is clearly incorrect. Real dividend growth has lagged behind real GDP per capita growth in all but one country, averaging just –0.1%, and the correlation between the two is -0.30.’ Higher economic growth is correlated with lower dividends and vice versa.

Conclusion

The clearly established negative correlation between economic growth and share price growth has a number of implications:

• It must be understood that correlation by itself does not establish causality - i.e. the negative correlation between share price growth and return on shares and GDP per capita growth does not establish whether rapid economic growth causes lower returns on shares, or higher returns on shares causes lower GDP growth, or some other factor(s) causes both. The negative correlation however does exclude that high share prices cause high GDP per capita growth, and therefore that economic policy should aim to have high share price growth in order to stimulate economic growth. The negative correlation shows that high returns on shares are either irrelevant (i.e. non-causal) for economic growth or if causal then higher returns on shares are negative for economic growth.
• Underperformance of shares in emerging economies with high growth rates would be expected to flow from the negative correlation between GDP per capital growth and share prices. Other things being equal, the more rapid the growth of GDP per capita the greater the underperformance of shares that would be anticipated - this clearly fits China.
• As the negative correlation of GDP per capita growth and share prices exists in both developed and developing economies pointing to underperformance of share prices in emerging markets does not justify the claim that underperformance of shares in such markets is due to government policy – except that such an underperformance would necessarily follow from the (desirable) policy of ensuring high GDP per capita growth.
• If share prices are negatively correlated with GDP per capita growth, but dividend payments are positively correlated with returns on shares, this again fits the case of China. China experiences high GDP per capita growth but low dividend payments due in substantial part to the large role played in the economy by State Owned Enterprises which pay extremely low, or zero, dividends to the state.
• Once again correlation does not establish causality, but the negative correlation of share price growth with per capita GDP growth, while dividend payments are positively correlated with share prices, means either that low dividend payments are beneficial for economic growth, that economic growth creates low dividend payments, or some other factor(s) causes both trends.
• Given the rate of growth of living standards is positively correlated with GDP per capita growth, whereas share prices are negatively correlated with GDP per capita growth, there is no reason to assume a positive correlation between rising living standards and rising share prices - i.e. that rising share prices will enhance living standards. Given that the aim is rising living standards there is therefore no reason for governments to target high share prices.

Taking the specific case of China, the negative correlation between share prices and GDP per capita growth is one reason this blog, which has a positive analysis of the growth of China’s economy, has a negative position as regards any major expectations for China’s share market.

Evidently the Financial Times and John-Paul Smith dealt primarily with short term trends in share prices which can can be affected by very different factors to the long term correlations outlined above. However:

• It is misleading to present underperformance of shares in emerging markets as being due to negative government policies when underperformance of shares in rapidly growing developing economies would be precisely expected to be correlated with their rapid growth.
• The implications of the negative correlation between GDP per capita growth and share prices, in both developed and developing economies, are far wider than those raised by John-Paul Smith and in the Financial Times article.

Notes & References

A 2nd wave of bank share price Armageddons?

The current falls in the prices of bank shares are widely covered in the financial media. However the overall pattern and scale of the decline in market evaluation of banking shares is still not sufficiently frequently drawn together. The aim of this note is therefore to show this pattern in clear terms. It reveals market estimations of a second wave of major bank share price collapses. This is particularly significant as it follows a lack of recovery of the first wave - i.e. there is a clear pattern of a spreading pool of 'collapsed' share prices of major banks. The cases of Barclays, Bank of America and Société Générale in this wave will be looked at in particular in comparison with the earlier cases of Citigroup, RBS and Lloyd's Banking Group.

The 1^st wave of bank share price collapses

The overall features of the wave of bank collapses in 2008 is too well known to need rehearsing here. Attention is simply drawn to the fact that, in addition to the formal bankruptcy or takeover of investment banks (Lehman Brothers, Bear Stearns, Merrill Lynch), a series of very large retail banks were left in a state which, from the point of view of their share prices, means they may be regarded as ‘collapsed banks’ – ‘collapsed’ for present purposes being defined as having suffered a fall of more than 90% in share price. These share prices may also be regarded as 'collapsed' in the sense that no significant recovery has taken place.

Attention is drawn to these features as there is sometimes an implicit assumption that share prices in these banks may have recovered at least to some degree in line with the overall recovery in share prices. This is clearly not the case.

Figure 1

 

Figure 1 therefore shows the share prices for three of the largest such banks – Citigroup, RBS and Lloyd’s Banking Group. The share prices in all these banks rapidly fell by around 90% during the financial crisis – the falls beginning, as a leading indicator of the coming crisis, in all cases between December 2006 (Citigroup) and February 2007 (RBS and Lloyd’s Banking Group). To make possible international comparisons, as with other banks considered here, share prices have been converted into dollars using daily exchange rates.

Figure 1 shows clearly the 1^st wave of collapses in bank share prices and that no recovery has taken place from it. On 9 September 2011 the shares of Citigroup had fallen by 94.8% from their peak, the shares of Lloyd’s Banking Group had fallen 95.9% and the shares of RBS had fallen 97.6%

The 2nd wave of bank share price collapses

The significance of the new current wave of severe bank share price falls is illustrated in Figure 2. This shows that the share prices of a further series of very large banks – Barclays, Bank of America and Société Générale are analysed – had initially suffered falls in share prices during the financial crisis which were as large as the first group of ‘collapsed’ bank share prices. However unlike the first group this second set of institutions had undergone a genuine recovery in share price. What is now significant, therefore, is that the recent share prices of this second group are now falling sharply downwards towards the level of ‘collapsed’ banks – i.e. a second wave of very large banks are being pulled into a maelstrom of  share price collapses.

Figure 2

 

Specifically, for the banks considered above:

• Barclays share price, after peaking in February 2007, by 23 January 2009 had fallen by 95.5%. By 17 September 2009, however, Barclays share price had recovered to to 59.3% below its peak. By 9 September 2011 Barclays share price had again fallen to 85.3% below its peak level – i.e. a second wave of severe share price falls had set in.
• Bank of America’s share price, having peaked in October 2007, by 6 March 2009 had dropped by 93.4%. It then recovered to 61.1% below its peak by 14 October 2009. By 9 September 2011 however its share price had again fallen to 85.3% below its peak.
• Société Générale’s shares, having peaked in May 2007, by 9 March 2009 had fallen by 88.3%. They then recovered to 58.8% below their peak by 29 September 2009. By 9 September 2011 they had fallen to 87.9% below their peak.

The share price pattern of these banks is therefore common. They reached low points early in 2009 before recovering by autumn 2009. Since then at first a slow, and now rapid, renewed decline in their share price has taken place. In other words the market judgement is that a new wave of severe share drops of major banks – including Barclays, Bank of America and Société Générale – is taking place down towards the level of the 1st wave of ‘collapsed’ (in terms of share prices) banks, that is Citigroup, RBS and Lloyd’s.

In order to show this pattern Figure 3 shows the share prices of both the 1st and 2nd wave of share price falls in the banks considered.

Figure 3

 

Severe price falls, by normal standards, are of course taking place for most other banks – including Goldman Sachs, Deutsche Bank, HSBC etc – but these cannot be said to be approaching the level of near 90% falls for the banks analysed above.

Conclusion

The conclusion of the above trends is evident. Bank share prices show a clear pattern of two waves of ‘collapse’ among large retail, or retail/investment, banks.

• The 1st wave saw falls of 90% plus in the shares of banks such as Citigroup, Lloyd’s and RBS. No significant recovery in their share prices has taken place since – which is why they may, in terms of share prices, be characterised as ‘collapsed’ banks.
• A number of other large banks – of which Barclays, Bank of America and Société Générale may be taken as examples - initially suffered very severe share price falls during the international financial crisis but then saw a significant recovery. Their share prices are, however, now again falling towards the level of the 1st group of banks. It, of course, remains to be seen whether this trend will continue, but a current tendency of a 2nd wave of banks being drawn into crises which essentially destroys share prices is clear.

The market therefore clearly judges that the pool of ‘collapsed’ banks, in terms of share prices, is threating to spread in a second wave.

US share price decline continues to match fall after 1929

The latest market shifts confirm that to date in the present financial crisis the fall in US share prices continues to be at a post-1929 rate – see Figure 1.

Figure 1

Last Friday, 27 February, was the 352nd trading day since the all time high of the Dow Jones Industrial Average on 9 October 2007. Since its peak the Dow has declined by 50.1%. On the 352nd trading day following its 1929 high point the Dow had declined by 55.6%. The difference between the falls is within the range of short term fluctuations. Overall the rate of decline of US share prices since October 2007 is tracking the 1929 descent.

As may be seen from Figure 2 the current decline of share prices is far more rapid and deep than either of the other two major falls in the 20th century – that following the oil price increase of 1973 or following the implosion of the dot com share bubble in 2000. These trends confirm again that the only relevant scale of comparison for the current decline of US share prices is with 1929 itself.

Figure 2

The only fundamental difference between the current decline and that of 1929, so far, is the duration of the fall. After 1929 it took the Dow Jones 713 trading days to reach its bottom – the trough being on 8 July 1932 by which time the Dow it had lost 89.2% of its value. It remains to be seen for how long the current decline will continue. However although the duration of the drop is at least as yet not as great as after 1929 it is as rapid.

While the financial collapse is therefore genuinely on a scale to be compared to 1929, the overall declines registered in the productive economy are still far smaller than the post-1929 falls. However the declines in the productive economy have just commenced and data is still coming in. Some drops, such as in trade, are approaching scales seen post-1929. However the decline in US GDP in the 4th quarter, now revised downwards to an annualised decline of 6.2 per cent, is about two thirds of the 9.4 per cent drop in US GDP seen between 1929 and 1930 – however the rate of decline is accelerating. While quite sufficient information is now in to make clear that the current financial crisis is only comparable to 1929, far exceeding the shifts seen in a normal recession, further data is still required to see whether it will be meaningful to compared the fall in production to that after 1929 or not.

Comparison of the current decline in share prices with 1929

In a post in October this blog warned against any underestimation of how severe and how long a fall in share prices could be. The current post provides further comparative information on this, updates relevant data, and justifies the view that a relevant comparison for the current fall in share prices is with 1929 rather than a normal cyclical recession.

Taking first the latest data, Figure 1 shows the daily movement of the Dow Jones Industrial Average following its recent peak on 9 October 2007 with the movement following its peak on 3 September 1929. It updates the data given in the earlier post on Key Trends in Globalisation.

As may be seen the present decline in the Dow continues to be entirely comparable in magnitude, at this stage of the crisis, to its fall in 1929-32.

Figure 1

 

In order to show that such a severe decline in nominal share prices is a specific feature of the 1929 and 2007 crises, and not typical of any recession, Figure 2 shows a similar graph for the four most serious declines in the Dow in the last century - those starting in 1929, 1973, 2000, and 2007.

For the three earlier declines the data covers the period from the peak price preceding the decline to its low point. The data for the decline starting in 2007 are up to the latest available date - the close of trading on 20 November 2008.

Figure 2

 

It may be seen that the falls in nominal prices starting in 1973, associated with the oil price increases and recession of that year, and in 2000, following the bursting of the dot com financial bubble, were far less severe than the drops in either 1929 or in 2007.

The fall in real share prices following 1973 is understated by this graph, as at that time inflation was higher than in 1929, 2000 or 2007, while the decline in real terms following 1929 is somewhat exaggerated as at that time the overall price level in the economy was falling. But the differences of order of magnitude are sufficient to make the pattern clear.

The fall in nominal share prices following both 1929 and 2007 far exceeds that of any other drop in the last century. From the angle of share prices it is entirely justified, and without exaggeration, to speak of the present crisis as comparable only to 1929.

The difference between the fall starting in 2007 and that in 1929 is only, at present, in the duration of the decline. The decline after 1929 continued for 712 trading days before reaching its bottom on 7 July 1932. The decline following the peak of 9 October 2007 has so far continued for 284 trading days - slightly under forty per cent of the period of the decline following 1929.

From this data it should not be implied that is being argued that the current speed of decline of US share prices, which has no parallel since 1929, will continue for as prolonged a period as in 1929. The point is merely being made that the current decline in US share prices is not comparable to a simple cyclical recession - even of a severe type. The only standard of comparison for the current fall in US share prices is with 1929, and the experience of 1929 indicates that a fall in share prices can go significantly further than it has so far.

Taking one measure of fundamentals, the ratio of total equity, including private equity, to GDP this measure has dropped from 1.8 in 2000 at the height of the dot com boom, to 0.8 at present. But during the trough of the decline in share prices in the mid-1970s it was only 0.4. An equivalent drop would imply a further 50 per cent fall of the Dow to around 4000 - meaning that the total fall from peak to trough would be around 72 per cent.

Such decline is not inconceivable in a very serious fall in share prices - the nominal decline of the Dow after 1929 was 89 per cent although, as overall price deflation was occurring, the decline in real prices was somewhat less. The maximum decline of the Nikkei since its peak at the end of 1989 has been 80.2 per cent - although as Japan has been undergoing moderate price deflation the fall in terms of real prices is again slightly less. A fall of the Dow by more than seventy per cent would not, therefore, be historically unprecedented - which is not the same as to say it will occur. At the close of trading on 20 November the decline of the Dow since its peak in 2007 was 46.7 per cent.

Regarding the potential period of a decline of share prices it took 25 years after 1929 for the Dow to regain its pre-crash level even in nominal terms. As Figure 3 shows the performance of the Nikkei after 1989 was even worse than that of the Dow after 1929 - the Nikkei was still registering new lows almost 18 years after the beginning of its decline.

The figures given above are for major stock exchanges, those of the US and Japan, during major stock market crashes. Performance of more minor stock markets, considered historically and for earlier periods, can be worse.

Prices on the French stock market failed even to keep up with inflation for 53 years from 1900-1952. Prices on the German stock market failed to keep up with inflation for 55 years from 1900-1954. Prices on the Japanese stock market, in addition to the recent fall of the Nikkei, failed to keep up with inflation for 51 years from 1900-1950. Prices on the Italian stock market failed to keep up with inflation for 73 years from 1906-1978.

The reasons for making these points and comparisons are not merely theoretical. Private individuals and companies may believe that the decline in share prices will necessarily only be short term and will be followed by a bounce and take investment decisions accordingly. Governments, when undertaking major programmes of purchases of shares in banks and other institutions, may believe that they are buying 'at the bottom of the market' and this will be followed by the taxpayer making profits as share prices rise.

There is no justification for such views. Historical comparisons show that there remains considerable space for downward potential in share prices even in nominal terms. Historical study indicates it may take many years, possible decades, for share prices to recover to previous levels from falls of the present magnitude.

How bad can a fall in share prices be?

Frequent comparisons are made between the current financial crisis and that of 1929. In one aspect this is misleading. In 1929 there was not merely a financial crash but an extremely severe downturn of production. At present, given the dynamism of the Asian economies, there is no reason to believe that the world economy as a whole will see a severe decline in output - or indeed that the world as a whole, as opposed to some individual countries, will suffer any decline in output at all.

This, however, does not mean that the financial depression of share markets cannot potentially be very severe. This is particularly relevant to the various plans now being unveiled for governments to purchase shares in banks. Indeed it is disturbing that one of the erroneous aspects of media coverage of these proposals is that they frequently implicitly assumes that bank share prices in the future must rise.[1] This leads to wrong evaluation of risk.

If bank shares must inevitably rise, after a period of falls that has already taken place, then there is evidently no significant risk for the taxpayer in buying them. If, however, bank shares may fall further, and remain depressed for a prolonged period, then the risk is great. It is therefore illustrative to make a comparison to risk following periods of exceptional financial turmoil created by asset price bubbles.

Figure 1 therefore shows the percentage decline in the Dow Jones Industrial Average from its maximum value in 1929, on 3 September, to its minimum value in the decline following this - on 8 July 1932. The fall of the Dow was 89.2 per cent in a little under three years.

Figure 1

   

Also shown for comparison in Figure 1 is the current percentage decline of the Dow since its high on 3 October 2007 until the end of trading on Friday 10 October 2008. As may be seen the decline after October 2007 initially was slower than in 1929. However, the decline accelerated rapidly last week. Last Friday was the 255th trading day since the 2007 peak of the Dow and by that point the Dow had fallen by 40.3 per cent. In comparison 255 trading days after the peak in 1929 the Dow had declined by 35.7 per cent. That is, in the current decline, the fall of the Dow since 2007 is slightly greater than following 1929.

What marked the historic scale of the 1929 fall, however, was not only the severity of the decline but its duration. By 255 trading days after the peak of 1929 the Dow had not even fallen by half the amount that it was eventually to drop. After 255 days of trading following the 1929 peak the Dow was down 35.7 per cent, while the eventual fall was to be 89.2 per cent. That is, 255 days into the decline what looked like a severe fall was, in fact, more towards the beginning of the decline than the end.

Figure 2 shows the overall historical pattern of the Dow post 1929 until its eventual recovery. It took until 23 November 1954 for the Dow to rise to its 1929 peak level in nominal terms - as inflation occurred in the intervening 25 years the real level of the Dow in 1954 was, of course, still below its level at the peak in 1929. It would have taken 25 years, even in nominal terms, for an investment made at the peak of the market in 1929 to have been recovered.

Figure 2

 

However, as some significant decline in share prices has already taken place in the current cycle, assume that an investment had been made on the 255th day of the decline of the Dow after 1929 - that is on 10 September 1930. The Dow did not recover to that level until 15 January 1951. That is, it would have taken 21 years for the original investment to have recovered to the same level even in nominal terms - and longer in real terms.

It may be argued, however, that 1929 was a wholly exceptional event and therefore such a pattern could not possibly be repeated. Apart from the fact that the present crisis is the worst since 1929, and one should therefore not make any such assumptions, experience of other countries shows that repetition of comparable patterns is quite possible.

Figure 3 shows the movement of Japan's Nikkei 225 share index since it reached its peak on 29 December 1989. The maximum fall of the Nikkei after this was only marginally less than that for the Dow following 1929. The maximum percentage fall of the Nikkei so far, following its peak, was 80.5 per cent - compared to an 89.2 per cent decline for the Dow following 1929.

The bottom, so far, of the Nikkei following this peak was reached on 28 April 2003 - that is the Nikkei continued to fall for almost 14 years after its peak.

Again, to make a comparison to the current situation, the 255th day of trading of the Nikkei after the peak was on 17 January 1991. By that time it had declined by 39.7 per cent - entirely comparable to the 35.7 per cent fall of the Dow on the 255th day of trading after its peak in 1929 and the 40.3 per cent fall of the Dow on the 255th day of trading after the 2007 peak.

Again, to make comparisons, this means that on this comparable day of the decline after its peak the majority of the fall in the Nikkei was still to come.

Figure 3

 

What is striking about the Nikkei pattern, however, is two further trends.

First, it may not be the case that the bottom has yet been reached. As may be seen, by the end of trading on 10 October 2008 the Nikkei was again moving downwards towards its minimum level - on 10 October the Nikkei was 78.7 per cent below its 1989 peak compared to the maximum fall recorded so far of 80.5 per cent.

Second, and most important for present purposes, the length of the depression of the Nikkei is even longer than that for the Dow following 1929.

10th October 2008 was the 4,624th trading day after the peak of the Nikkei in 1989. On the 4,624th trading day on the Dow after the Great Crash the Dow was 55.0 per cent below its peak. In comparison the Nikkei after the same interval was 78.7 per cent below its peak. That is, in terms of length of the decline, the depression of the Nikkei has so far been more severe than the decline of the Dow after 1929.

This comparison shows clearly that the post-1929 fall of the Dow is not at all an unparalleled event following the collapse of a major asset price bubble. On the contrary Japan is living through a share price collapse that is of entirely comparable dimensions.

Again a comparison to the 255th day of trading in the current cycle, that is the situation on 10 October 2008 compared to the Dow's peak in October 2007, is revealing. By the 255th day of trading after the peak the Nikkei had fallen 39.7 per cent. The eventual maximum fall was to be 80.5 per cent. That is 255 days into the fall not even half of the eventual decline had taken place.

What conclusions follow from this? First, it should be stressed that what has been looked at above is the movement of share indices, not of the shares of individual companies. No mechanical comparisons should be drawn - bank shares have fallen more than the Dow or FTSE 100 for example so a greater part of the decline might be behind. But, equally, it is entirely possible for the value of shares of an individual company to in effect fall to zero - as occurred with AIG, Lehman Brothers, Bradford and Bingley, and Northern Rock - whereas, even at their worst, it is entirely implausible that a share index will lose all of its value, that the value of all quoted companies falls to zero.

The downside risk on individual shares, may be seen by considering the arithmetic of the proposals announced today for the British government to purchase shares in the British banks Royal Bank of Scotland (RBS), HBOS and Lloyd's TSB.

Alistair Darling, the British Chancellor of the Exchequer (finance minister) announced on radio, just after 8pm today, that the government would purchase RBS shares at 65.5p. The government has also stated it will purchase HBOS shares at 113.6p. It is, therefore, entirely possible that share prices in RBS and HBOS may decline further, even to zero, and the tax payer will lose that investment - that is, major taxpayer risk has been assumed.

What is notable about the case of RBS is that, as the private sector was unwilling to subscribe the necessary capital at 65.5p a share, the private sector, that is the market, rates the shares of RBS as worth less than 65.5p a share. The government is therefore forcing the taxpayer to undertake an investment which the market itself is not willing to take - i.e. the government is taking an investment position riskier than the market. The potential for loss in such a position is evident - and strongly illustrated by the fact that those who invested in the earlier rights issue of RBS this year had lost £8 billion by 10 October 2008.

What is the consequence of assuming this risk? The 'taxpayer' is not an abstract entity but consists not only of individuals but of profitable, that is viable, companies. Therefore the risk the UK government has taken has been put onto the shoulders of individuals and viable companies, instead of it being concentrated on RBS, HBOS, and Lloyds TSB shareholders. This is damaging to other viable companies and to individuals.

While it would be extremely unwise to judge such a fundamental issue by short term market movements nevertheless the scale of the risk was illustrated even on the morning of 13 October - following the British government announcement. The RBS share price fell to as low as 49.6p by 11 45 am - a price that would have entailed large losses for the taxpayer on any shares purchased at 65.5p. HBOS was trading at an even lower level, falling to a low of 83p – that is 27 per cent below the price at which the government proposed to buy shares.

The British government, of course, had to state that it was ready to step in to take over RBS, HBOS or Lloyd's TSB to ensure their orderly functioning and guarantee deposits - as it did with Northern Rock or Bradford and Bingley. But that would occur in conditions in which, in essence, the market concluded that the value of the shares was zero – in which case zero would have been a fair price.

This loss would have course been borne by RBS, HBOS, or Lloyd's TSB shareholders. But losses would have been focussed on them - in accord with market operations. Instead, now, the potential for large loss for individuals and viable companies (that is taxpayers) has been assumed by buying shares at prices which are judged by the market to be above their value.

This also illustrates the mechanism by which viable companies and individuals are placed at risk. The injection of capital by the government will necessarily raise RBS share prices - not necessarily eventually compared to the 65.5p share purchase price by the government but compared to the real eventual market value (which may be as low as zero). Existing shareholders will therefore be able to sell at above value prices while the government is unable to sell its stake - suffering severe potential losses for taxpayers if the price is below 65.5p. Taxpayers would thereby lose while existing shareholders would gain.

The government, evidently, hopes bank shares will rise and this loss for the taxpayer, that is viable companies and individuals, will not occur. It may be fortunate. But there is no justification for risking taxpayers money in purchasing bank shares at what are above the market values if such an intervention had not taken place. This is to assume great risk for which viable companies and individuals may suffer.

The data given above on movements of share prices after the collapse of major asset bubbles clearly gives no justification for a belief that shares have necessarily reached their bottom and the only potential is up and not down. Comparison to previous falls in share prices after the collapse of asset price bubbles, on the contrary, shows there is considerable potential for further losses. If banks can raise private capital they should, of course, do so - as with Barclays and HSBC. However the assumption of considerable risk to individuals and viable companies by the government in purchasing shares in RBS, HBOS and Lloyds TSB, rather than standing entirely willing to take over the orderly and guaranteed running of these companies if they prove unviable, is therefore not justified.

Notes

[1] To take one example the Observer’s editorial on Sunday morning stated: 'the banks should not treat the government's equity stake like a simple loan. They cannot expect that, when the current crisis has passed, government will step back from its investment without extracting a profit. Having part-nationalised the banks, the state must manage its shareholding to yield the best return for the taxpayer'.

The fall in asset prices is causing the liquidity crisis, not the liquidity crisis the fall in asset prices

The announcement that the US government is to follow the example of the British government, and buy shares in US banks in order to recapitalise them, brings to the fore one of the most important issues in the present financial crisis -one which has major financial and policy implications.

The international financial crisis is at present focused in two areas – asset values (share prices, house prices etc) and liquidity (drying up of interbank lending leading to a paralysis of the financial system). But the analysis of the interrelation of these two processes is vital in understanding how to tackle them.

The decisive question is whether it is the fall in asset values that is driving the liquidity crisis, or it is the liquidity crisis that is driving the fall in asset values? Different measures, with hugely different financial implications, follow from the two different answers.

The real cause of the crisis is that assets in the US are overvalued – in the end due to the overvaluation of the dollar. That is, if real market process were allowed to develop, and in the end no one will be able to stop them from operating in some form, these assets will be shown to have less value than their former and present prices.[1]

As these assets devalue down to their competitive market values this weakens, or renders insolvent, the balance sheet of institutions directly or indirectly holding them – that is they suffer loss or bankruptcy.

As these assets fall in value this necessarily produces a liquidity crisis - as financial institutions cannot lend, due to their overstretched balance sheet, nor are they willing to lend to other banks who are also greatly overstrained and may therefore not be able to repay loans. That is, the fall in asset values drives the liquidity crisis.

The present financial crisis logically started in the weakest and most overvalued part of US assets – the sub-prime mortgage market. But it is not confined to them and therefore spreads through other parts of the financial and asset system – the crash of share prices on Wall Street during the last week being the latest manifestation of this.

The policy conclusion that flows from this is that as the assets were overvalued, therefore someone will inevitably suffer loss as a result of their decline - this is unfortunately unavoidable. The only economic question is who will suffer this loss?

The aim of policy must be to ensure that this loss is concentrated as narrowly as possible on the source - that is those institutions most economically responsible for the crisis. If this is not done losses will necessarily spread through the system and those not responsible (viable firms, taxpayers etc) will suffer higher than necessary losses. In this case economic rationality coincides with morality.

This means shareholders in financial institutions which took the wrong decisions, that is who bought/created such overvalued assets, should not receive funds from those who were not responsible. If shareholders in institutions which made wrong decisions are safeguarded then others who bear no responsibility for this failure - viable companies, taxpayers, depositors etc - will suffer corresponding losses. 

The government must naturally be prepared to step in to ensure the functioning of the banking system, but it should not be bailing out bank shareholders. This model was used, for example, in the rescue of Swedish banks in the 1990s and was also carried out in the nationalisation of Fannie Mae, Freddie Mac, and AIG in the US and Northern Rock and Bradford and Bingley in the UK. This will involve the state taking over these failing banks to ensure the functioning of the banking system.

Liquidity in this situation will be maintained by two steps. First, immediately, central banks must substitute for the frozen lending markets by themselves lending or using existing nationalised banks to do so – Northern Rock can be used in the UK for this purpose. Second, after nationalisation of the insolvent parts of the banking system, these banks can recommence interbank lending. This route will minimise losses for ordinary depositors, tax payers, consumers, and viable companies.

However consider the policy implications if, on the contrary, it is believed that it is the liquidity crisis which is driving the fall in asset values. In this case the first step is the same  – the money markets are flooded with liquidity. But the second step is totally different and will result in losses being spread through the system to taxpayers and viable companies.

If it is believed that it is the liquidity crisis which is driving the fall in asset values then, as the liquidity crisis is overcome through injections of money, assets would rise in value. In particular bank shares will rise in value. It is, therefore, rational for taxpayers money to be put into bank shares, that is to ‘recapitalise’ the banks - indeed it is conceived this may even lead to a profit in the medium term. This is what the UK government proposed and it is now announced the US will follow this. However consider what will occur if this analysis has got the dominant direction of causation wrong – that is that it is the fall in asset prices which is driving the liquidity crisis and not vice versa.

First, despite the liquidity injection, the money put into bank shares will not halt the fall in asset values as this is not being driven by liquidity problems but by a quite other mechanism. Falls in asset values will therefore continue – that is the taxpayer will suffer big losses on the sums put into bank shares (this may be worsened by existing shareholders selling shares at values that have been temporarily artificially inflated by the capital injections). These losses will then make it much harder to indemnify deposit holders, tax payers, maintain public spending, assist viable companies etc – in other words those responsible for the crisis will haved been safeguarded at the expense of those not responsible for the crisis and losses will be spread through the system.

Second, the interbank lending market will have a strong tendency to jam again or to remain jammed – because the downward pressure on asset values means that banks will not be able or willing to start large scale lending. It is quite probable that in the end, to unjam this situation, the banks will have to be nationalised anyway but by this time large amounts of the taxpayers money that had been put in to 'recapitalise’ the banks will have been lost.

The mistake of believing that it is the liquidity crisis that is driving the fall in asset prices, rather than understanding that it is the fall in asset prices that is driving the liquidity crisis, will therefore lead to major losses for the taxpayers and viable companies and also fail to resolve the banking crisis.

To look at the real world, it is quite clear from the facts that it is in reality the fall in asset values that is driving the liquidity crisis and not vice versa. The crisis started in asset values, in sub prime mortgages. This, in turn, revealed that other assets, in the first places the shares and other holdings of US financial institutions, were themselves overvalued. It was this that then created the freezing of interbank lending and the other aspects of the liquidity crisis. In short it is the fall in asset prices that created the liquidity crisis, not vice versa.

Governments will not succeed in overcoming the present situation until they analyse correctly its dominant direction of causation and adopt the appropriate policies.

Notes

[1] This working out of market forces could theoretically take many forms. One, which did not occur, was a gradual devaluation of the dollar over a prolonged period of years. A second would be a sudden drop in the exchange rate of the dollar - that is nominal dollar prices might not fall but their real international value would be cut. A third would be that the exchange rate of the dollar remained the same but that the nominal dollar price of assets fell sharply. But in all cases the price of the dollar denominated assets will fall. Which of these variants occurs depends on other, more short term, factors. In the present phase the dominant process is the third.