Last week, the Australian Financial Review was doing its best to spruik the ongoing prospects for the Australian share market in their front page article “Cashed-up funds have $70bn to invest”. The article is only available online to subscribers, but this quotation sums it up:
analysts cite the volume of cash stockpiled as a reason for stocks to keep rising
Mostly consisting of quotations from people in the equity business (who all arguably stand to benefit from talking up the market), the authors do include some data to support the proposition as well:
The latest data released by the Australian Bureau of Statistics shows that fund managers have increased their cash holdings to about 18 per cent of the $880 billion they manage, or about $160 billion. If managers were to return their cash holdings to more normal levels, there would be about $70 billion available for investment, with the local sharemarket receiving up to $30 billion.
The image of a wall of cash on the sidelines waiting to spill over into equity markets is compelling, but does it make sense? The power of this commonly used image arises from the idea that cash is somehow transforming into shares, when of course for every buyer there is a seller who gets the cash, so share trading never changes the total amount of cash in the system (note that aggregate money supply can change through central bank operations and banking deposit creation, but that is a whole other story beyond the sharemarket and is not part of the standard “cash on the sidelines” argument). Of course, this does not stop share prices from going up or down.
So, if the cash in the system does not change, what is going on?
For a start, most commentators who consider cash on the sidelines to be a driver of stock prices look at the percentage of total assets held in cash; this post on the Market Thoughts blog is a good example. The problem with this approach is that, while it creates an apparent correlation, it is trivial and has no predictive power. This is because a fall in the value of shares (or indeed any risky asset class) will reduce the aggregate value of investment assets and, with the total amount of cash staying constant, the percentage in cash automatically increases. It will then decrease again when shares increase in value again. But of course, this means that the move in cash percentages is coincident with movements in share prices and provides no better prediction than simply saying “shares have fallen and eventually they will go back up”.
A more sophisticated interpretation is to say that “cash on the sidelines” does not mean aggregate cash in the system, but refers specifically to cash held by a segment of the market, namely investors in money market funds (the figure most often measured). The argument could go something like this: many investors target asset allocation mixes and if their cash holdings become too high, they would be forced to buy more shares, thereby adding to demand for shares, shifting the demand curve and pushing prices higher. I have a few problems with this argument.
First, if this phenomenon is significant, why would the cash imbalance have arisen in the first place? Presumably because investors became more risk averse and demand for shares decreased. If so, I would argue that what would push share prices up is sentiment: investors become less risk averse and increase their demand for shares once more. This may, as a consequence, push the cash in the segment back down if investors’ net buying was from participants outside their segment. If so, I would see that the change in cash is a consequence of a change in demand, not the cause.
Second, by segmenting you need to consider not only the demand side, but the supply side. Investors cannot simply trade among themselves, otherwise the cash in their segment would not change. Their increased holding of cash would have to be offset by a reduced holding in cash by other segments, such as the corporate segment, which may affect the supply curve. A plausible scenario would be that corporates, starved of cash, begin issuing shares to raise cash, possibly at a discount to market prices. In the process, they channel the cash back from the investor segment and push share prices down. While I am not saying this would happen, I don’t think it is much less plausible than the explanation that excess cash shifts the demand curve for investors.
A related point is that the segment of investors with money in money market funds is bigger than asset allocators. While asset allocators may feel the pressure of their cash holdings increasing relative to their target, other investors may be becoming more risk averse, offsetting the effect of the asset allocators.
So much for theoretical arguments. Since this is the Stubborn Mule, I really should look at what the data says. The AFR article refers to Australian Bureau of Statistics data on the cash holdings of fund managers. Looking at the history of cash and deposits held by fund managers and comparing it to the Australian All Ordinaries share price index, there is no obvious pattern of increases in cash holdings leading to rises in share prices.
Fund Managers’ Cash and the Share Market (1988-2009)
In case there is a hidden pattern in these time series charts, we can also look at a plot of quarterly changes in the share price index against changes in cash holdings of fund managers.
There are a few outliers in this chart, but certainly nothing to support the claim that cash on the sidelines can push the market up. Of course, it may be that there is a lag before the effect on share prices takes effect, so here is a plot where changes in the All Ordinaries index are lagged by six months.
Cash versus Lagged Share Prices (1988-2009)
Even with the lag, there are a couple of occasions where an increase in cash was followed by share market rises, but the biggest cash build-up was followed by share market declines and for more modest changes in cash levels, the change in the share price index is distributed across gains and losses.
So, neither theory nor data supports the cash on the sidelines argument. And finally, I’ll just note if you believed in the argument back when the article I linked to above was published, you would have gone long stocks in June 2008! So, the next time you hear someone talking about all the “cash on the sidelines”, tell them they are talking nonsense.
UPDATE: Commenting on this post, James suggested calculating the correlation between changes in cash balances and changes in share prices for a variety of time lags. The chart below takes up this suggestion. The correlations are quite variable, which is consistent with my contention that build-ups of cash do not drive share prices. But it is also interesting to note that, if there is a pattern to be found, it is that the correlations are mostly negative, suggesting that increases in cash are more often followed by share price declines not rallies.
Correlation of Cash and Lagged Share Prices (1988-2009)
I should concede that this chart, which stretches out five years (20 quarters), ends up looking for correlations far beyond where even the most hopeful “cash on the sidelines” advocates would expect to see a relationship.