Hello Bogleheads,
I've been reading up on levered etf strategies and am very interested in allocating a small but sizable portion of my net worth in them. I am thinking allocating my risk towards HedgeFundie's 55/45 UPRO/TMF strategy. However, I have some reservations about the TMF leg of this strategy, given the higher than expected inflation prints we've been seeing recently, and higher than expected pace of rate hikes and the damage that might do to TMF.
I'm curious if anyone has read this paper by Michael Gayed, Leverage for the Long Run. It seems like a managed approach to just UPRO would be superior in this market. Has anyone considered my line of thinking and have done any data analysis to weigh the pros/cons of the two strategies? Thanks
UPRO/TMF or Managed UPRO?
Re: UPRO/TMF or Managed UPRO?
I skimmed through it but I don't think that I read it before. I think I've seen it discussed before though.zerofighter2148 wrote: ↑Sun Jan 16, 2022 7:59 am Hello Bogleheads,
I've been reading up on levered etf strategies and am very interested in allocating a small but sizable portion of my net worth in them. I am thinking allocating my risk towards HedgeFundie's 55/45 UPRO/TMF strategy. However, I have some reservations about the TMF leg of this strategy, given the higher than expected inflation prints we've been seeing recently, and higher than expected pace of rate hikes and the damage that might do to TMF.
I'm curious if anyone has read this paper by Michael Gayed, Leverage for the Long Run. It seems like a managed approach to just UPRO would be superior in this market. Has anyone considered my line of thinking and have done any data analysis to weigh the pros/cons of the two strategies? Thanks
The TL;DR is that (i) there is a relationship between volatility and whether the market is above or below a moving average, (ii) performance is worse when volatility is high, and (iii) a timing approach based on the moving average (risk on for the LETF when above the moving average and risk off to treasuries when below the moving average) has historically provided better returns, both on absolute and risk-adjusted basis.
I personally don't like risk-on/risk-off approaches that slam completely in and out, especially ones based on moving averages. I have never been able to develop a statistically predictive model based on moving averages that improves returns; timing just doesn't seem to work consistently.
With that said, I think that there is some value in using estimates of future volatility to condition the allocation, using something like target volatility or risk parity approaches. Future volatility can be estimated with some statistical power, without the indirect moving-average analog. I've done some testing of adaptive allocation based on volatility approaches related to the HFEA in this thread. My conclusion is that you can reduce portfolio volatility significantly, making it easier to stick with the approach, although over sufficiently long periods portfolio returns are not necessarily improved.
The most direct analog to Gayed's proposed approach would be to use target volatility (for example, here with cash as out of market asset). This is comparable to leveraging a balanced fund, which is essentially a 60/40 HFEA.
It's hard to say what the behavior would be like going forward; TMF will likely offer better crash protection than cash but worse drag on performance. Who knows which will win out.
-
- Posts: 16054
- Joined: Fri Nov 06, 2020 12:41 pm
Re: UPRO/TMF or Managed UPRO?
It might be a good strategy. Basically instead of TMF, you'd get out and hold cash during the downturns.
The only gotcha is that you might have to pay STCG depending on the timing of bear markets breaching 200D MA.
The only gotcha is that you might have to pay STCG depending on the timing of bear markets breaching 200D MA.
Re: UPRO/TMF or Managed UPRO?
There is already an ETF that does this.zerofighter2148 wrote: ↑Sun Jan 16, 2022 7:59 am It seems like a managed approach to just UPRO would be superior in this market.
Its 20% UPRO and moves in and out of cash.
https://www.howardcmetfs.com/funds/defender500
-
- Posts: 58
- Joined: Sat Oct 09, 2021 8:29 pm
Re: UPRO/TMF or Managed UPRO?
If you are worried about the interest rate effects on TMF, you can use TYD(3x IEF) or the new TYA(2.5X IEF). https://www.portfoliovisualizer.com/tes ... odWeight=0
Re: UPRO/TMF or Managed UPRO?
I would assume NTSX would end up being far more tax-efficient though, correct?Booogle wrote: ↑Sun Jan 16, 2022 3:59 pmThere is already an ETF that does this.zerofighter2148 wrote: ↑Sun Jan 16, 2022 7:59 am It seems like a managed approach to just UPRO would be superior in this market.
Its 20% UPRO and moves in and out of cash.
https://www.howardcmetfs.com/funds/defender500
Re: UPRO/TMF or Managed UPRO?
Actually LGH is more tax efficient:DanFFA wrote: ↑Sun Jan 16, 2022 6:36 pmI would assume NTSX would end up being far more tax-efficient though, correct?Booogle wrote: ↑Sun Jan 16, 2022 3:59 pmThere is already an ETF that does this.zerofighter2148 wrote: ↑Sun Jan 16, 2022 7:59 am It seems like a managed approach to just UPRO would be superior in this market.
Its 20% UPRO and moves in and out of cash.
https://www.howardcmetfs.com/funds/defender500
https://screener.fidelity.com/ftgw/etf/ ... ymbols=lgh
https://screener.fidelity.com/ftgw/etf/ ... mbols=ntsx
-
- Posts: 224
- Joined: Sat Apr 25, 2020 8:28 am
Re: UPRO/TMF or Managed UPRO?
Plenty of people have 'considered' your thinking - that paper has been shared all over the internet.zerofighter2148 wrote: ↑Sun Jan 16, 2022 7:59 am Hello Bogleheads,
I've been reading up on levered etf strategies and am very interested in allocating a small but sizable portion of my net worth in them. I am thinking allocating my risk towards HedgeFundie's 55/45 UPRO/TMF strategy. However, I have some reservations about the TMF leg of this strategy, given the higher than expected inflation prints we've been seeing recently, and higher than expected pace of rate hikes and the damage that might do to TMF.
I'm curious if anyone has read this paper by Michael Gayed, Leverage for the Long Run. It seems like a managed approach to just UPRO would be superior in this market. Has anyone considered my line of thinking and have done any data analysis to weigh the pros/cons of the two strategies? Thanks
The problem I have with a timing strategy like that is that it's not predictive - it's a lagging indicator. You only learn that the market is drawing down AFTER the fact, not before. In any case, a 200DMA isn't even a great moving average and it issues too many false negatives.
There are predictive timing algorithms out there such as using other timing strategies that I won't really get into here that work much better. They're predictive, not reactive unlike a moving average strategy. I've modelled a moving average strategy all the way to the 1920s and my results weren't exactly as good as the paper above (which I will also point out isn't peer-reviewed nor is it an academic paper).
I'm not a big believer in momentum strategies myself so I wouldn't dissuade anyone from investing like the above but I personally wouldn't recommend it.