Ron Bewley says it’s too soon to tell what effect the global financial crisis has had on properly-constructed, long-term gearing strategies.
People only ask questions such as, “Does gearing still work?” when the market is bad – and that is not the time to assess a long-term strategy. We, and others, have argued that the investment horizon for geared equities should be at least five years. That means we cannot completely assess the impact on gearing strategies of the fall from the November 2007 high until at least November 2012. No one knows where the market will be by then. This case study is a “best endeavours” attempt to assess, in the interim, the impact of the global financial crisis (GFC) on a few gearing strategies.
We have used daily data on the S&P/ASX 200 price and accumulation indexes to separate out daily capital growth and dividend payouts since January 2001 – the date when the RBA first published indicative margin loan interest rates. We assume dividends are 80 per cent franked and that the tax rate is 46.5 per cent throughout. For simplicity we assume after-tax dividends are re-invested daily, unless there is a shortfall, when dividend income is less than interest. This “loss” is carried over until there is a surplus of dividend income over interest cost. We ignore capital gains tax as we assume that the strategy is for at least five years, so we want to see what will be “rolled over” to the next period.
Here we consider only simple strategies to illustrate a few points:
a) No gearing.
b) One-off gearing: an initial 50 per cent loan-to-value ratio (LVR) with a margin call at 80 per cent, which would be sold down to 70 per cent in the event of a call.
c) Constant gearing: daily adjustments to maintain a 50 per cent LVR.
d) Managed gearing: an initial 50 per cent LVR with a “pre-emptive” sell-down at a 65 per cent LVR (to 60 per cent) and a “topping up” at a 40 per cent LVR, back to 50 per cent.
There are 949 separate starting points, since we close the experiment at September 30, 2009. In the upper panel of the table we show the median (of the 949 annualised returns), the best and the worst. Of course, three descriptors of the whole distribution of returns doesn’t tell the whole story – but that is all we have space for. In no five-year period did an ungeared strategy produce a negative return – and that’s after tax is paid on the dividends. The last “buy” day was October 1, 2005. Surprisingly, the downside was not much greater for one-off gearing (-0.2 per cent per annum versus 1.6 per cent per annum) but the median and upside are far better.
Constant gearing is largely impractical on a daily basis, but we note that if we had invested that way, the distribution of returns is more spread-out but slightly tilted towards better returns. The managed gearing strategy that we specified proved to be unsuccessful. Of course, loan management rules that take other information – such as the state of the market – into account could be very different. Managing LVRs is, however, critical for risk management. The major problem with all empirical comparisons such as this is that the results are highly dependent on the timing of the end-point – in this case, in the recovery phase after a major bear market. To give some insight, we now add some “behaviour” into the experiment. For any five-year term ending on or after October 1, 2007 (just after the Lehman Brothers collapse) we extend the investment period to end on September 30, 2009 (and we still annualise the results).
This behaviour is tantamount to saying that people hang on when times are bad. Of course, if times get even better, the results will favour hanging on even more. The results from “hanging on” are shown in the bottom panel of the table. It has no impact on the median and the best, as hanging on is designed only to improve the bad results. Stretching the time-frame only a little improves the worst ungeared result from 1.6 per cent per annum to 8.8 per cent per annum! The improvements are even greater in the geared strategies. Three conclusions follow from this study:
1) One-off gearing works well and doing several of these together – as in regular gearing – has a lot going for it.
2) Gearing out of a bear market is critical to success if the investor geared into it.
3) Ad hoc management of LVRs might not prove fruitful for returns but is critical in managing risk. Of course, how those investors would have gone if they had bought at the top will not be known for another three years.
