HEDGEFUNDIE's excellent adventure Part II: The next journey

[oldcomputerguy]

Several contentious posts were removed. See General Etiquette:

We expect this forum to be a place where people can feel comfortable asking questions and where debates and discussions are conducted in civil tones. Discussions are about issues, not people. If you disagree with an idea, go ahead and marshal all your forces against it. But do not confuse ideas with the person posting them. At all times we must conduct ourselves in a respectful manner to other posters. Attacks on individuals, insults, name calling, trolling, baiting or other attempts to sow dissension are not acceptable. If you feel that someone has attacked you or otherwise violated the policies of this forum, do not respond in kind. Instead, please click the report button on the offending post. This is the quickest method to notify the site moderators.

Please keep the discussion civil, and centered on the original topic.

[elderwise]

Question, for others who are more knowledgeable..

I get it that doing quarterly rebalancing will trigger STCG, but if doing annual (yeah cAGR may not be as good as quarterly or lookback monthly) but, wont that mean LTCG?

I am asking because I have close to 100K vested into this in taxable, if (hopefully) it goes upto 250K over next few years or maybe more or less.If I hold this 100K invested for more than a year, isnt that xYZ gain LTCG?? I know it will bump my AGI..but,, never the less just wanted to understand..

Thanks.

[MotoTrojan]

Question, for others who are more knowledgeable..

I get it that doing quarterly rebalancing will trigger STCG, but if doing annual (yeah cAGR may not be as good as quarterly or lookback monthly) but, wont that mean LTCG?

I am asking because I have close to 100K vested into this in taxable, if (hopefully) it goes upto 250K over next few years or maybe more or less.If I hold this 100K invested for more than a year, isnt that xYZ gain LTCG?? I know it will bump my AGI..but,, never the less just wanted to understand..

Thanks.

Yes, and annual actually does quite fine. It actually did better depending on starting date for the GFC, and overall from 1955-present (and 1982-present) it did best via the Simba spreadsheet, if you take black monday out of the mix.

If I were doing this in taxable I would certainly utilize annual to make it LTCG, but the tax-drag is still going to greatly impact your risk-adjusted return.

[BogleBobby]

Question, for others who are more knowledgeable..

I get it that doing quarterly rebalancing will trigger STCG, but if doing annual (yeah cAGR may not be as good as quarterly or lookback monthly) but, wont that mean LTCG?

I am asking because I have close to 100K vested into this in taxable, if (hopefully) it goes upto 250K over next few years or maybe more or less.If I hold this 100K invested for more than a year, isnt that xYZ gain LTCG?? I know it will bump my AGI..but,, never the less just wanted to understand..

Thanks.

If you do annual rebalancing, you can make sure that you also get LTCG rates.

However, one thing to consider is that with either FIFO accounting, or by choosing specific lots to sell, you can probably always get LTCG rates after being in this strategy for a year (even if you do quarterly rebalancing) since you'll generally always have shares in either account that you've had >1 year. In that instance, then you'll have to compare whether it's better to sell old shares with a lot of gains and get LTCG, or sell more recent shares which don't have as many gains but will get STCG. Or you may be able to share sells at a loss and not incur capital gains.

[schismal]

Question, for others who are more knowledgeable..

I get it that doing quarterly rebalancing will trigger STCG, but if doing annual (yeah cAGR may not be as good as quarterly or lookback monthly) but, wont that mean LTCG?

I am asking because I have close to 100K vested into this in taxable, if (hopefully) it goes upto 250K over next few years or maybe more or less.If I hold this 100K invested for more than a year, isnt that xYZ gain LTCG?? I know it will bump my AGI..but,, never the less just wanted to understand..

Thanks.

I have a few thousand in a taxable account. This is small enough to rebalance through contributions. If I had a larger position and was not rebalancing through contributions, I would not do anything less than annual unless/until you can guarantee LTCG rates on older positions. Definitely not in the first year.

FWIW I tried to estimate the tax consequences in this post.

[NMBob]

thanks for the tax estimates Schismal

[Lee_WSP]

Or you may be able to share sells at a loss and not incur capital gains.

No. The thirty day before and after rule sinks that idea. It's probably easier come tax time to sell lots with ltcg or zero stcg.

[DosCommas]

Or you may be able to share sells at a loss and not incur capital gains.

No. The thirty day before and after rule sinks that idea. It's probably easier come tax time to sell lots with ltcg or zero stcg.

It is possible to use SPXL as a TLH pair for UPRO to avoid wash sales. Since the start of the adventure there have been a few opportunities to tax loss harvest already given the volatility. There does not seem to be a suitable TLH pair for TMF at this time but hopefully something becomes available in the future.

[butricksaid]

Or you may be able to share sells at a loss and not incur capital gains.

No. The thirty day before and after rule sinks that idea. It's probably easier come tax time to sell lots with ltcg or zero stcg.

No it wouldn't... assuming that you're rebalancing quarterly (not monthly) AND not making additional contributions then for each rebalancing event you are either:

1) Selling TMF to buy UPRO
2) Selling UPRO to buy TMF

In no scenario are you re-buying the same asset within 30 days before/after for this type of tax loss harvesting.

[BogleBobby]

Or you may be able to share sells at a loss and not incur capital gains.

No. The thirty day before and after rule sinks that idea. It's probably easier come tax time to sell lots with ltcg or zero stcg.

No it wouldn't... assuming that you're rebalancing quarterly (not monthly) AND not making additional contributions then for each rebalancing event you are either:

1) Selling TMF to buy UPRO
2) Selling UPRO to buy TMF

In no scenario are you re-buying the same asset within 30 days before/after for this type of tax loss harvesting.

I think Lee was thinking that I was suggesting selling shares of the under-represented asset for TLH. (e.g. I need to add more UPRO, so I sell losses of UPRO for a loss and buy them back immediately.) But I wasn't suggesting that. You have the scenario correct that I was suggesting.

However, as DosCommas pointed out, you could potentially sell your under-represented asset for TLH if you swapped it for an almost equivalent asset (e.g. sell UPRO to harvest losses and buy SPXL).

[BogleBobby]

One nice property of this strategy is that it doesn't rely on a bull bond market.

For instance, from 2000-2018:
- TMF had a return of 10.8%
- 40/60 UPRO/TMF strategy had a return 13.3%
- 55/45 UPRO/TMF strategy had a return 12.3%
- SPY had a return of 4.5%

But even if you reduce the TMF return during that period to something much lower, such as 5%, by lowering the return of each year by a consistent amount (and thus preserving the sequence of returns), then you still get:

- (Reduced) TMF return of 5%
- 40/60 UPRO/TMF strategy with a return 9.8%
- 55/45 UPRO/TMF strategy had a return 9.6%

So even without a great bond return, you'd still beat the SPY return of 4.5%. As has been stated in this thread before, the negative correlation between LTTs and SPY have been a big part of the success of this strategy.

If you look at historical correlation between LTTs and stocks, you can see that they have been negatively correlated recently:



If you look further back than the graph above, you can calculate that LTTs and stocks were actually negatively correlated in the 1960s.

Does anyone know exactly when the fed started managing interest rates differently? I think I've read that it may be the 1980s, but you can see that stocks and bonds were actually positively correlated for most of the 1980s and 1990s.

[Lee_WSP]

After Volker I believe.

[BogleBobby]

Thanks. That sounds right. After re-reading the OP in part I, HF showed the correlation of LTTs and stocks from May 1986 to now, so it must've been around that time.

One of the nuances pointed out in the original OP was that LTTs and Equities are most strongly negatively correlated when equities have a really large drawdown. Since the 1990s didn't have any "really bad years" for the SP500, LTTs and equities didn't have any events to cause them to be negatively correlated.

You can see that on this chart of SPY returns since 1987 and the five-year correlation between LTTs and equities.



If you go farther back, the relationship isn't really there, but there also weren't a lot of "really bad years" to evaluate.



Looks like we had "really bad years" in 1973 and 1974 and LTTs didn't do much to offset them.

[MotoTrojan]

One nice property of this strategy is that it doesn't rely on a bull bond market.

For instance, from 2000-2018:
- TMF had a return of 10.8%
- 40/60 UPRO/TMF strategy had a return 13.3%
- 55/45 UPRO/TMF strategy had a return 12.3%
- SPY had a return of 4.5%

But even if you reduce the TMF return during that period to something much lower, such as 5%, by lowering the return of each year by a consistent amount (and thus preserving the sequence of returns), then you still get:

- (Reduced) TMF return of 5%
- 40/60 UPRO/TMF strategy with a return 9.8%
- 55/45 UPRO/TMF strategy had a return 9.6%

4.5% X 3 X 0.4 + 5% *0.6 = 8.4%

4.5% X 3 X 0.55 + 5% *0.45 = 9.675%

Not seeing much rebalancing bonus here, at-least in the 55/45 variation. Of course it relied on a bond bull market to achieve the incredible returns since 1982; without it, you should expect low single digit CAGR improvement, unless rates rise and you trail the S&P500. Low single digit CAGR improvement is still huge, hence why I am using it .

[HawkeyePierce]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

[BogleBobby]

I was reading a bit into why LTT bond funds didn't experience very good returns in the mid 1970s (while stocks were falling) and it appears that the federal reserve was interested in lowering rates to promote economic growth, but they also had to fight inflation.

The U.S. economy, which was already struggling, fell into recession in November 1973. Shocked by double-digit inflation figures, central bankers hiked rates for five months in March 1974, introducing a monetary policy known as "stop-go," in which the Federal Reserve would alternate between fighting inflation by raising interest rates and then trying to revive economic growth by cutting them again

https://www.businessinsider.com/every-i ... uly-1974-2

[BogleBobby]

One nice property of this strategy is that it doesn't rely on a bull bond market.

For instance, from 2000-2018:
- TMF had a return of 10.8%
- 40/60 UPRO/TMF strategy had a return 13.3%
- 55/45 UPRO/TMF strategy had a return 12.3%
- SPY had a return of 4.5%

But even if you reduce the TMF return during that period to something much lower, such as 5%, by lowering the return of each year by a consistent amount (and thus preserving the sequence of returns), then you still get:

- (Reduced) TMF return of 5%
- 40/60 UPRO/TMF strategy with a return 9.8%
- 55/45 UPRO/TMF strategy had a return 9.6%

4.5% X 3 X 0.4 + 5% *0.6 = 8.4%

4.5% X 3 X 0.55 + 5% *0.45 = 9.675%

Not seeing much rebalancing bonus here, at-least in the 55/45 variation. Of course it relied on a bond bull market to achieve the incredible returns since 1982; without it, you should expect low single digit CAGR improvement, unless rates rise and you trail the S&P500. Low single digit CAGR improvement is still huge, hence why I am using it .

UPRO return was actually -1.5% during that period (because of the big drawdowns in 2000/2001/2002 and 2008), so I think you need to do the calc using the UPRO return (rather than SPY*3) to include the effect of volatility decay, management fees, and borrowing costs. Sorry for not including that in my original post.

[BogleBobby]

Also, just a little more context on the interest rates in the 1970s:

A key to understanding pretax real interest rates in the last quarter century is recognizing that real interest rates in much of the 1970s were extraordinarily low owing largely to the two OPEC oil shocks, which decreased investment demand and increased world saving by transferring wealth from the high-consuming developed countries to OPEC.

https://www.bostonfed.org/-/media/Docum ... wp91_6.pdf

Some researchers regard the 1973 "oil price shock" and the accompanying 1973–74 stock market crash as the first discrete event since the Great Depression to have a persistent effect on the US economy.[41]

The embargo had a negative influence on the US economy by causing immediate demands to address the threats to U.S. energy security.[42] On an international level, the price increases changed competitive positions in many industries, such as automobiles. Macroeconomic problems consisted of both inflationary and deflationary impacts.

https://en.wikipedia.org/wiki/1973_oil_crisis

[MotoTrojan]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

What is the overall allocation you went with? I keep vacillating on what to do with the $10K I left in the 55/45; leave it, merge it with the EDV allocation, or have extra fun (and make more work) and use it for target volatility or some other cute timing method.

[MotoTrojan]

One nice property of this strategy is that it doesn't rely on a bull bond market.

For instance, from 2000-2018:
- TMF had a return of 10.8%
- 40/60 UPRO/TMF strategy had a return 13.3%
- 55/45 UPRO/TMF strategy had a return 12.3%
- SPY had a return of 4.5%

But even if you reduce the TMF return during that period to something much lower, such as 5%, by lowering the return of each year by a consistent amount (and thus preserving the sequence of returns), then you still get:

- (Reduced) TMF return of 5%
- 40/60 UPRO/TMF strategy with a return 9.8%
- 55/45 UPRO/TMF strategy had a return 9.6%

4.5% X 3 X 0.4 + 5% *0.6 = 8.4%

4.5% X 3 X 0.55 + 5% *0.45 = 9.675%

Not seeing much rebalancing bonus here, at-least in the 55/45 variation. Of course it relied on a bond bull market to achieve the incredible returns since 1982; without it, you should expect low single digit CAGR improvement, unless rates rise and you trail the S&P500. Low single digit CAGR improvement is still huge, hence why I am using it .

UPRO return was actually -1.5% during that period (because of the big drawdowns in 2000/2001/2002 and 2008), so I think you need to do the calc using the UPRO return (rather than SPY*3) to include the effect of volatility decay, management fees, and borrowing costs. Sorry for not including that in my original post.

That is fair, although as MoneyMarathon pointed out UPRO (or TMF) rebalanced with even a cashzero (zero return asset) will reduce the volatility decay.

[HawkeyePierce]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

What is the overall allocation you went with? I keep vacillating on what to do with the $10K I left in the 55/45; leave it, merge it with the EDV allocation, or have extra fun (and make more work) and use it for target volatility or some other cute timing method.

I copied yours.

[MotoTrojan]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

What is the overall allocation you went with? I keep vacillating on what to do with the $10K I left in the 55/45; leave it, merge it with the EDV allocation, or have extra fun (and make more work) and use it for target volatility or some other cute timing method.

I copied yours.

If you don't mind me asking, is this your only LETF exposure? % of portfolio? Are you counting it as part of your US equity allocation?

I am including the EDV variant as part of my Total US/Large allocation, but not the 55/45 adventure.

[HawkeyePierce]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

What is the overall allocation you went with? I keep vacillating on what to do with the $10K I left in the 55/45; leave it, merge it with the EDV allocation, or have extra fun (and make more work) and use it for target volatility or some other cute timing method.

I copied yours.

If you don't mind me asking, is this your only LETF exposure? % of portfolio? Are you counting it as part of your US equity allocation?

I am including the EDV variant as part of my Total US/Large allocation, but not the 55/45 adventure.

I'm now using NTSX as my core holding in the rest of my portfolio so I still have leveraged Treasuries but that fund uses a range of durations, not just LTT. I count my UPRO/EDV "adventure" separately from the rest of my portfolio, which is now 100/20. My thinking is that I'll get the risk/return exposure of a 100% equities portfolio while also capturing the ballast effect of a bond allocation.

A 32% allocation to NTSX gives me a 19.2% allocation to bonds through its Treasury futures plus a 28.8% allocation to US large-caps.

If I count it all as one portfolio it leaves me with a ~25% allocation to Treasuries. Haven't done the math but at a glance I think it works out to a weighted average duration of around 15 years.

[jaj2276]

Big changes to two of my slices today. Both my 21-day vol lookback and 16-vol target slices have swung to major TMF overweight. After today's rebalance, my entire portfolio (which consists of a 50% 40/60 that hasn't been rebalanced since end of June and is currently at 33/67) will be 33% UPRO, 67% TMF. So overall, a slight overweight to TMF compared to the OG 40/60 for September.

[BogleBobby]

That is fair, although as MoneyMarathon pointed out UPRO (or TMF) rebalanced with even a cashzero (zero return asset) will reduce the volatility decay.

True. I don't think there's an easy formula to calculate the rebalancing effect when looking at leveraged products.

[MotoTrojan]

That is fair, although as MoneyMarathon pointed out UPRO (or TMF) rebalanced with even a cashzero (zero return asset) will reduce the volatility decay.

True. I don't think there's an easy formula to calculate the rebalancing effect when looking at leveraged products.

It would be interesting to see how much of a rebalancing bonus there was from unleveraged S&P500 and LTT from 1982-present, vs. just outright LTT performance boost.

[caklim00]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

Me too. Four 25% Pieces of Pie, just submitted on M1 the changes, went ahead with the Quartely change submissions today to get everything inline
1) 43/57 UPRO/EDV Quarterly
2) 55/45 UPRO/TMF Quarterly
3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

[BogleBobby]

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

What are these two methods exactly? I think #3 is a 20-day lookback period looking at the volatility of each asset and then weighting each asset inverse to their volatility. If that's right, what is #4?

[Walkure]

Today marks my second quarterly rebalance (using new money in taxable.)
Stats as of this morning's trading window:
The 40/60 allocation has drifted to 36/64.
Cumulative UPRO: 14%
Cumulative TMF: 83%
Cumulative Total MWRR: 53.03%

[MotoTrojan]

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

What are these two methods exactly? I think #3 is a 20-day lookback period looking at the volatility of each asset and then weighting each asset inverse to their volatility. If that's right, what is #4?

Target volatility. Basically the same as #3 but you ONLY look at UPRO's volatility and adjust the allocation to target a specific vol (in this case 20% was used). Then whatever allocation is leftover you use TMF.

This has a lot more research behind it than risk-parity look-back. Works great with unleveraged funds too. Market timing of course...

[MotoTrojan]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

Me too. Four 25% Pieces of Pie, just submitted on M1 the changes, went ahead with the Quartely change submissions today to get everything inline
1) 43/57 UPRO/EDV Quarterly
2) 55/45 UPRO/TMF Quarterly
3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

Interesting. Are you hedging your bets or just like to see how they all compare for fun?

[entangledquanta]

Can someone outline the mechanics of these two allocations?

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

I would like to know how the two allocations were made?

[MotoTrojan]

Can someone outline the mechanics of these two allocations?

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

I would like to know how the two allocations were made?

Read the posts near the end of the Part I thread, this has been discussed significantly. Portfolio Visualizer has tools for backtesting both the risk parity look-back (which compares the 2 assets) and the target volatility (which only compares the one asset to a user-input target).

https://www.portfoliovisualizer.com/tes ... ndowSize=1

The allocations change frequently, most people are using 20 day lookbacks (monthly).

[Hydromod]

Can someone outline the mechanics of these two allocations?

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

I would like to know how the two allocations were made?

Read the posts near the end of the Part I thread, this has been discussed significantly. Portfolio Visualizer has tools for backtesting both the risk parity look-back (which compares the 2 assets) and the target volatility (which only compares the one asset to a user-input target).

https://www.portfoliovisualizer.com/tes ... ndowSize=1

The allocations change frequently, most people are using 20 day lookbacks (monthly).

The allocations bounce around much less with 60-day lookbacks. It's not clear that rebalancing with the larger changes due to the shorter lookbacks would have actually bought much performance.

[Xile]

Curious if anyone has experimented with blending in like 10% of BTAL into this strat. It excels in volatile markets and typically rises when stocks fall while not dragging too hard in up markets. I was thinking maybe 50 upro/40 TMF/10 BTAL as a starting point. Modeling it in MPI Stylus shows a slight reduction in performance in up periods obviously, but a noted increase in performance in down periods most of the time. Biggest problem is not being able to lookback beyond 2011.

[caklim00]

I just joined MotoTrojan on the EDV train. I was swayed by the lower total ER and path-dependency risks of TMF.

Me too. Four 25% Pieces of Pie, just submitted on M1 the changes, went ahead with the Quartely change submissions today to get everything inline
1) 43/57 UPRO/EDV Quarterly
2) 55/45 UPRO/TMF Quarterly
3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

Interesting. Are you hedging your bets or just like to see how they all compare for fun?

No caps or anything, just whatever PV spits back for me. This is such small part of my portfolio (Started 30k up to 32.67k now) that even if target volatility puts UPRO at 100% for 3 straight months I can live with it. Even though this is small for me I'd be thinking about it too much if I went all in on target volatility.

This is the nicest thing about M1, it is so easy to do this.

[caklim00]

Can someone outline the mechanics of these two allocations?

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

I would like to know how the two allocations were made?

Here are PV links for both
3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
https://www.portfoliovisualizer.com/tes ... 0&total1=0

4) 20 volatility 30/70 UPRO/TMF Monthly
https://www.portfoliovisualizer.com/tes ... total1=100

Note: for 20% volatility I use the last 20 days but I weight the most recent days more. So days 1-10 more than days 11-15 more than days 16-20. Intuitively it makes more since to me that the most recent week should probably be weighted higher than 3-4 weeks ago. It worked out better historically better this way but might not in the future. Volatility on UPRO went down slightly recently so its 30% UPRO instead of 28% UPRO if you weight days 1-20 equally.

[caklim00]

Can someone outline the mechanics of these two allocations?

3) 20 day lookback for risk parity 41/59 UPRO/TMF Monthly
4) 20 volatility 30/70 UPRO/TMF Monthly

I would like to know how the two allocations were made?

Read the posts near the end of the Part I thread, this has been discussed significantly. Portfolio Visualizer has tools for backtesting both the risk parity look-back (which compares the 2 assets) and the target volatility (which only compares the one asset to a user-input target).

https://www.portfoliovisualizer.com/tes ... ndowSize=1

The allocations change frequently, most people are using 20 day lookbacks (monthly).

The allocations bounce around much less with 60-day lookbacks. It's not clear that rebalancing with the larger changes due to the shorter lookbacks would have actually bought much performance.

I looked at your analysis but just playing around with PV seemed to have better results with shorter periods. That said, at a certain point this becomes too much work at increasingly shorter periods. Who knows if it will pay off but I have a bit in each of the strategies just to see.

[Forester]

What would have been the annual return in 2008 for a 20% vol target?

[MotoTrojan]

What would have been the annual return in 2008 for a 20% vol target?

I can’t recall the years return but drawdown was only 40% or so which helped a lot compared to 55/45’s deep >60% drawdown; since the low both performed similarly.

[BogleBobby]

One other bond fund that I've considered is the ITT-3X fund in Simba's workbook (which I believe is equivalent to the TYD ETF).



It doesn't get you the high annual returns that TMF does, but it also performs better when rates are rising. It does well when backtesting back to 1955 because it doesn't have the same losses during the 1960s and 1970s that TMF had, and still has done pretty well recently to offset stock losses recently.

[Lee_WSP]

I brought up the 3x ITT fund before and was shown that the actual performance was not even close to 3x and therefore not currently worthwhile.

[mrwheelerdealer]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

[Forester]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

60% US
25% Developed / Europe
10% Emerging
5% GDXJ

Leveraged ETFs might be too volatile for asset class switching, so own the world, with a US tilt?

[mrwheelerdealer]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

60% US
25% Developed / Europe
10% Emerging
5% GDXJ


Leveraged ETFs might be too volatile for asset class switching, so own the world, with a US tilt?

Is there any proof from a backtest that asset class rebalancing would be too volatile? Theoretically it should reduce risk, as factor investing is supposed to improve the Sharpe ratio?

Also, for your recommendation of world diversification, what would be recommended ETFs for implemenatation?

[MotoTrojan]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

Volatility hurts daily rebalancing leveraged funds so small-cap would likely have a big drag. I looked into international via DZK but it’s been shown to have a ~3% drag on top of the nominal index (simulated data). You could use an unleveraged EM fund perhaps to get some less correlated volatility.

[Lee_WSP]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

Volatility hurts daily rebalancing leveraged funds so small-cap would likely have a big drag. I looked into international via DZK but it’s been shown to have a ~3% drag on top of the nominal index (simulated data). You could use an unleveraged EM fund perhaps to get some less correlated volatility.

UPRO has had a significantly better 10 year run than TNA. In other words, "what he said".

[Kevin M]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

Volatility hurts daily rebalancing leveraged funds so small-cap would likely have a big drag. I looked into international via DZK but it’s been shown to have a ~3% drag on top of the nominal index (simulated data). You could use an unleveraged EM fund perhaps to get some less correlated volatility.

UPRO has had a significantly better 10 year run than TNA. In other words, "what he said".

Interesting. For July 2009 - Aug 2019:

1x:
SPY: 14.29%
IWM: 12.69%

3x:
UPRO: 35.70%
TNA: 24.04%

It's more obvious by looking at cumulative returns, but although small cap lagged large cap, the difference in the 3X versions is huge by comparison.

Kevin

[HEDGEFUNDIE]

This idea is actually intriguing to me.

One thing I don’t understand is why the equity portfolio is dominated by only market beta (and the S&P 500 at that!). Would it not make sense to diversify across geography and other risk factors, such as international, small cap, and/or value (are there ETFs that provide *3 factor exposure? I know there are many for international). Perhaps even adding a gold ETF would increase the ability of the portfolio to survive a worst case scenario?

Volatility hurts daily rebalancing leveraged funds so small-cap would likely have a big drag. I looked into international via DZK but it’s been shown to have a ~3% drag on top of the nominal index (simulated data). You could use an unleveraged EM fund perhaps to get some less correlated volatility.

UPRO has had a significantly better 10 year run than TNA. In other words, "what he said".

Interesting. For July 2009 - Aug 2019:

1x:
SPY: 14.29%
IWM: 12.69%

3x:
UPRO: 35.70%
TNA: 24.04%

It's more obvious by looking at cumulative returns, but although small cap lagged large cap, the difference in the 3X versions is huge by comparison.

Kevin

It’s really no surprise at all. Here are the StdDevs over that period:

IWM: 17.4%
SPY: 12.6%

Good ol’ volatility decay in action.

[Forester]

'Direxion Launches 3x Leveraged ETFs Tracking FTSE Developed Europe Index'
http://www.direxion.com/press-release/d ... rope-index

What's going on? https://www.portfoliovisualizer.com/bac ... 0&total3=0