I prefer this approach (use of gearing) to that of boosting returns by including more risky assets in my portfolio because, theoretically, using gearing can provide more increase in return for an increased level of risk than just including a larger proportion of high-risk, high-return assets in your portfolio. Apparently it has something to do with the tangent of the line from the interest rate on the borrowed funds to the point on the efficient frontier corresponding to your asset mix having a steeper slope compared to just moving along the efficient frontier... Don't ask me to explain! As far as I can tell it looks something like the diagram below:
The idea is that you can increase the return of your "efficient" portfolio more by taking on extra risk through the use of gearing (ie. along the red line) than by changing the asset mix to include more risky assets (moving along the efficient frontier, assuming you have the right asset mix).
I don't worry too much about the details, as the entire concept of the efficient frontier seems fairly academic. In practice, you don't have precise data about all the assets in your portfolio (historic standard deviation, return, and covariance for all of the assets in your portfolio) or for determining the efficient frontier (such data on all possible assets mixes). Also, the usual problem of predicting the future using historic data applies. In practice, it seems that a sensible, well-diversified asset mix lies pretty close to the efficient frontier, and the use of gearing will be preferable to the inclusion of more risky assets in your portfolio (provided the interest rate on your investment loan isn't too high).
The other benefit of using gearing is that by having a larger total investment you are more able to make use of diversification.
Several things about the use of gearing (borrowing to invest) seem important to me:
1. You have to manage your risk - it is stupid (very risky) to borrow to invest in just one particular stock, or sector (eg. tech, oil, bio etc). I was amazed to learn that around 15% of margin loans are for accounts that have only one stock in them! The chances are you are not the next Warren Buffet (I know I'm not), so it's much more sensible to diversify - either via a portfolio of at least 10-12 stocks (not all in the same sector), or else via an index fund. Or a combination of both.
2. You have to borrow at the best possible interest rate - over time I've opened new margin lending accounts with lenders that offer a lower interest rate and that allow you to trade on your margin account via an online discount broker. Interest rates may be lower if you borrow a larger amount. I also have a margin lending account with the bank that provided my home and investment property loans - as a preferred customer I get 0.25% knocked of the standard interest rate. Most recently I used a "portfolio loan" secured against the equity in my real estate assets (similar to a HELOC?) to borrow at the standard home loan rate and invest in a portfolio of US shares (see my posts about my "Little Book" portfolio).
3. You should borrow less than the maximum so that you have a buffer against getting a margin call when the next market correction/bear market/crash happens. If the average margin for the stocks in your margin lending account is 70%, then you probably shouldn't get your loan-to-value ratio (LVR) above 50%.
The table below shows what % drop in the market would trigger a margin call for different starting values of gearing:
Average 70% margin
Margin Call occurs if loan > 105% of margin value
% drop to initial initial initial
Capital Loan Total get a call LVR MU DER gearing
$100,000 $ 50,000 $150,000 54.6% 33.3% 47.6% 50.0% 1.5
$100,000 $ 60,000 $160,000 48.9% 37.5% 53.6% 60.0% 1.6
$100,000 $ 70,000 $170,000 43.9% 41.2% 58.8% 70.0% 1.7
$100,000 $ 80,000 $180,000 39.5% 44.4% 63.5% 80.0% 1.8
$100,000 $ 90,000 $190,000 35.5% 47.4% 67.7% 90.0% 1.9
$100,000 $100,000 $200,000 31.9% 50.0% 71.4% 100.0% 2.0
$100,000 $110,000 $210,000 28.7% 52.4% 74.8% 110.0% 2.1
$100,000 $120,000 $220,000 25.7% 54.5% 77.9% 120.0% 2.2
$100,000 $130,000 $230,000 23.0% 56.5% 80.7% 130.0% 2.3
$100,000 $140,000 $240,000 20.6% 58.3% 83.3% 140.0% 2.4
$100,000 $150,000 $250,000 18.3% 60.0% 85.7% 150.0% 2.5
LVR = loan-to-value ration (amount borrowed / total value of portfolio)
MU = margin utilisation (amount borrowed / max. amount allowed)
DER = debt-to-equity ratio (amount borrowed / your equity)
Gearing is basically the "muplitplier effect" that will apply to any losses or gains made by the portfolio on your equity. For example, if the gearing used is 2.0, then if your portfolio went up 10% your equity will have increased 20%. Similarly, if the market "corrected" -15% causing your portfolio to drop -15%, your equity will have gone down by -30%.
As the loan amount stays constant, any drop in the value of your portfolio will decrease the margin value (eg. 70% of the portfolio value). This also means that your margin utilisation increases. Most lenders will give you a "margin call" if your margin utilisation exceeds 100%. (There's often a "buffer" of 5%, so you'd get a margin call if the MU was >105%). When you get a margin call you have to bring your MU back down to less than 100%. This can be done in several ways:
* put some cash into your account (ie. pay off some of the loan) - this is best
* put in some stocks (eg. if you had some unencumbered stocks you could add into the margin account) - this is second best, as only the margin value of the stock (say, 70%) counts.
* sell off some stocks - you may not want to sell at the price prevailing when you get a margin call.
The most interesting column in the table is the % drop required to trigger a margin call. Looking at daily closing price data for the Australian All-Ordinaries Index (similar to the S&P500), I calculated the maximum % drop ever experienced after buying on any particular date for the past ten years. This period doesn't include any once-in-a-lifetime crash like '87 or '29, but does include the major bear market of the early noughties. I then grouped all the maximum decreases into bands - for the 2,531 days studied (1996-2006) I counted how many times the maximum drop was 0-5%, how many times it was in the range 5-10 % and so on. The tabulated results are:
max drop count probability
0% - 5% 1,069 42.2%
5% - 10% 438 17.3%
10% - 15% 480 19.0%
15% - 20% 439 17.3%
20% - 25% 105 4.1%
25% - 30% 0 0.0%
30% - 35% 0 0.0%
35% - 40% 0 0.0%
>40% 0 0.0%
TOTAL 2,531 100.0%
This shows that, at least for this past ten year period, if you'd invested on any randomly selected day for had a small (4.1%) chance of experiencing a drop of 20-25% before things began to improve.
You had a much bigger chance (21.4%) of experiencing a drop of over 15%. And drops of less than 15% are quite common.
This emphasises why it is important not to use excessive gearing - you have a significant liklihood of getting a margin call if your LVR is much over 60% (remember, an 18.3% drop will trigger a margin call if your LVR was 60%). Using gearing conservatively (say less than 50% LVR) with a well-diversified portfolio reduces the chance of a margin call significantly.
My plan is to maintain an LVR of around 50%, giving me a gearing ratio of 2. Assuming I can borrow funds at around 7.5% on average, and my portfolio return averages around 9.0%, I would expect my return on equity (ROE) to be 9.0% + (9.0% - 7.5%) = 10.5%. Assuming inflation averages around 3.0%, this extra 1.5% return would mean my real (after inflation) return is increased by 33% using fairly modest gearing. And this would have a big impact on the final balance.