Long-term? No. I can see some possibility for very short-term speculation. Very long though and the defects of the position start to overwhelm the benefits of the leverage. One is almost always better off achieving the leverage another way. My experience with the folks who design these contracts is that they don't think carefully about who the natural longs & shorts would be. They think about trading jazz. Consequently, the economic design is often poor.
If you think about leverage in a CAPM framework for a minute to get the broad strokes of the risk-return tradeoffs, a leveraged contract or leveraged specific asset return will never produce a risk-return advantage because one can always lever the whole portfolio up or down. Levering a specific asset is always suboptimal. A levered specific asset literally can't outperform. If it did, they did the math incorrectly. Now that's an equilibrium result where everyone shares consensus expectations. But the broad strokes of that still apply. Unless I am speculating in only one thing, one specific contract or asset, I will always be better off levering the entire speculative position up or down not just one asset. That's part of why the actual application of leveraged contracts is so narrow. Non-institutional investors face tougher constraints on their use of leverage so they may, from time to time, find it preferable to get "bad" leverage through a specific contract rather than no leverage at all. But as I said, when I have done the head's up return comparisons between a levered contract and direct leverage, the indirect leverage has had a high transactions or opportunity cost.
The leverage obtained through the leveraged ETFs for example is not as effective as leverage obtained directly (i.e., by simply borrowing directly and investing in a long position) - at least the last time I did the math on the returns to the two positions. It's been so long since I did the head's up portfolio return comparison that I don't recall the details. Usually the differences in returns arise because of marking to market and resetting the leverage - just as a periodically rebalanced position will diverge from a true buy & hold strategy over a longer horizon than the rebalancing period. If I really had for example a ten year horizon and a specific target to go with that, I'd be best off matching the duration of my borrowing with my horizon rather than rolling short term debt. Creating the leverage directly one can do that. Embedded in a forward contract, one can't match durations. That's the kind of thing I mean. Depending on the investment purpose, that stuff can matter a lot. For an average investor engaging in ST speculation & depending on the quality of their forecasts, I'm not sure it matters.
I doubt many retail investors are going to be reading this however. Such concerns are usually in the domain of institutional investors. When I have done the head's up return comparison, I have been surprised at how poor the leverage effects were. Most individual investors would be better off using medium to long term borrowing against real assets as their leverage.
The leveraged ETF paper is not surprising. Its good to document it though; worth doing for it's own sake.
Glad you have a similar view to mine on LETFs. Do you think that LETFs have use cases in a long-term portfolio?
Long-term? No. I can see some possibility for very short-term speculation. Very long though and the defects of the position start to overwhelm the benefits of the leverage. One is almost always better off achieving the leverage another way. My experience with the folks who design these contracts is that they don't think carefully about who the natural longs & shorts would be. They think about trading jazz. Consequently, the economic design is often poor.
What defects would overwhelm the benefits? The paper shows that LETFs often outperform their traditional non-balancing leveraged portfolios.
Or are you saying that leverage in the long run won't help?
If you think about leverage in a CAPM framework for a minute to get the broad strokes of the risk-return tradeoffs, a leveraged contract or leveraged specific asset return will never produce a risk-return advantage because one can always lever the whole portfolio up or down. Levering a specific asset is always suboptimal. A levered specific asset literally can't outperform. If it did, they did the math incorrectly. Now that's an equilibrium result where everyone shares consensus expectations. But the broad strokes of that still apply. Unless I am speculating in only one thing, one specific contract or asset, I will always be better off levering the entire speculative position up or down not just one asset. That's part of why the actual application of leveraged contracts is so narrow. Non-institutional investors face tougher constraints on their use of leverage so they may, from time to time, find it preferable to get "bad" leverage through a specific contract rather than no leverage at all. But as I said, when I have done the head's up return comparisons between a levered contract and direct leverage, the indirect leverage has had a high transactions or opportunity cost.
The leverage obtained through the leveraged ETFs for example is not as effective as leverage obtained directly (i.e., by simply borrowing directly and investing in a long position) - at least the last time I did the math on the returns to the two positions. It's been so long since I did the head's up portfolio return comparison that I don't recall the details. Usually the differences in returns arise because of marking to market and resetting the leverage - just as a periodically rebalanced position will diverge from a true buy & hold strategy over a longer horizon than the rebalancing period. If I really had for example a ten year horizon and a specific target to go with that, I'd be best off matching the duration of my borrowing with my horizon rather than rolling short term debt. Creating the leverage directly one can do that. Embedded in a forward contract, one can't match durations. That's the kind of thing I mean. Depending on the investment purpose, that stuff can matter a lot. For an average investor engaging in ST speculation & depending on the quality of their forecasts, I'm not sure it matters.
That's totally fair. If you have a specific investment time horizon, using leverage that is obtained directly may be more effective.
I think the point that the paper is making is that leverage obtained directly (through borrowing) is quite correlated with LETF returns.
I think the value in this paper is towards retail investors, who do not have access to flexible and low borrowing costs through a prime broker.
I doubt many retail investors are going to be reading this however. Such concerns are usually in the domain of institutional investors. When I have done the head's up return comparison, I have been surprised at how poor the leverage effects were. Most individual investors would be better off using medium to long term borrowing against real assets as their leverage.