I’ve also reached the conclusion that I should use non-zero leverage, and MES futures look like the best option to achieve leverage. Does anyone have experience with buying-and-holding (quarterly rolling) MES futures, and not being classified as a professional investor by tax authorities?
1 Like
I wanted to update my thoughts about leveraged portfolios. Now when we are again having access to US ETFs after losing it temporary, I think I shouldn’t wait longer.
Some time ago I was thinking how to increase the amount of cash that I effectively borrow by using leveraged instruments. I was first thinking about buying another MES future, but the borrowed amount and potential losses were getting too big to let them run loose and expect them to be absorbed by a margin loan. So I started to think how much cash with respect to the nominal contract value should I keep: 30%? 1/3 ? 40%? And when and how to add cash to the margin reserve. Then I realized that I am actually trying to replicate leveraged ETFs and decided to take a closer look at them.
The advantages of leveraged ETFs vs. futures are:
- Granularity and flexibility. You borrow 70-100 USD per share of UPRO (no fractional shares) and can fine tune your total leverage.
- No need to calculate and keep cash reserves. Invest the whole cash amount available and be done with it.
Comes with costs, of course, but I think it’s worth it.
As a side note, I realized quite some time ago that portfolio simulations almost always assume a portfolio with an initial investment amount that is left to run without adding any new funds in it. Understandably, this is the easiest way. However I and many forum members, are in the accumulation stage, which has own peculiarities.
I aim at a portfolio of US stocks that has 75% of non-leveraged funds and 25% UPRO with 3x leverage, providing a total leverage of 1.5 - in US stocks portfolio at IB only. I did some quick simulations how this portfolio would act in the past in the accumulation mode. And I was convinced by the results and I had implemented it.
Now that I have discovered that UPRO has options expiring every Friday I will think how and if can/should I use options to harvest some premium from them 
.
9 Likes
Well, to make things clear from my side, I am not investing in UPRO only, it is just one piece of portfolio that provides the leverage to my desired level. SSO and UPRO have almost identical TER, so with UPRO I get more bang for my money.
Not exactly. The question they pose is which leverage is optimal, which is a genuinely interesting “research” question. Thank you for the info, for me it means that my 1.5 leveraging of the part of the total portfolio doesn’t look like excessive.
But then they seem to pretend that if the optimal leverage is 2, then you should invest in SSO only. And this is the part I challenge. Let’s say you decided to go with 2x leveraged S&P 500. You have at least 2 options
100% SSO, TER 0.92%
50% UPRO, 50% VOO, TER 0.5%.
Well, you know which forum you are in . Considering that 3x is the maximum leverage allowed for US ETFs, second allocation is probably the most cost effective possible one.
Have you considered NTSX?
Cheaper TER, notional exposure ~ 90/60 S&P500 Futures / Treasuries.
Probably a better bet in case of downturns 
2 Likes
I haven’t really checked the study, so may have missed critical methodology information, but yahoo finance usually doesn’t take dividends into account, they’re a price only database. The results of the study, which used them for data gathering, may (or may not) be affected by that.
1 Like
Great way to get leverage. Too bad there’s no NTSX combo with Swiss government bonds and MSCI World equity though
1 Like
There are NTSX, NTSI, and NTSE for US, ex-US, and EM, respectively. So getting some kind of world allocation should be easy.
Regarding government bonds I don’t know a clean and easy solution. Maybe hedge the bond portion to CHF manually?
2 Likes
I’m not sure the second allocation can stay around 2x without frequent rebalancing. This frequent rebalancing would have to stay below 0.42% cost.
EDIT: Checking numbers
If one accepts a maximum of 0.2 leverage deviation (coresponding to 60/40 or 40/60), then portfoliovisualizer.com gives me about one rebalancing per year in the last 5 years. A bit less frequently before, and 2 events in 2020. The rebalancing band was set to absolute 10%.
Rebalancing cost thus breaks even at a yearly:
0.42% / 10% = 4.2% = 2 x 2.1%
To spend more than 2.1% for buying or selling assets is highly unrealistic with a mustachian broker. IBKR wants 0.005 per share in FIXED (1 USD minimum). Assuming your share costs more than 5, then your cost is a maximum of 0.1%. Probably more like 0.01% or 0.001% looking at the actual price of UPRO and VOO.
That means my assumptions can be off by at least 21x, probably more like 210x, and still hold.
Congratulations, your solution seems efficient. 
1 Like
Note that I am accumulating, so I can reduce the need for selling to rebalance by directing new funds and dividends.
If someone speaks German perchance: I found “ZahlGrafs Exzellente Abenteuer” on Reddit.
ZahlGraf extensively discusses different leverages and allocations on S&P 500 and Treasuries. They also elaborate how they generate backtesting data down to 1943 including fund management cost and variable borrowing rates. The discussion has a focus on Germany (available instruments, taxes). The data is available for download (git clone).
If your German is not that good: You could probably use Google Translate and still get lots of information. But still there is heavy use of Reddit German (direct translations of English words. E.g. “mauerstrassenwetten” aka. “wallstreetbets”) which could confuse the translation engine (and normal people for that matter).
1 Like
Hi Helix,
Considering that CHF tends to be considered as a safe haven in case of a crash and typically the CHF becomes stronger vs USD is not holding simply CHF a good alternative to a government bond?
I‘m using a combination of VTI and UPRO to stay slightly leveraged into the US stock market. Aiming for 15% UPRO and 85% VTI. I plan to rebalance once my leverage ratio of 1.3 is below 1.2 or above 1.4. For example a 20% correction:
VTI goes from 85k to 68k, UPRO goes from 15k to 6k (simplified, I now). New leverage is 1.16. So I buy 5.1k UPRO and sell 5.1k VTI. Thus being back to a leverage ratio of 1.3 with 62.9k VTI and 11.1k UPRO. Market recovers back to previous levels (+25%), VTI goes from 62.9k to 78.6k and UPRO from 11.1k to 19.4k (again simplified). I‘m back at 98k and lost 2% due to volatility drag.
Is there a cheaper variant to stay leveraged? Buying call options? Futures?
1 Like
A cheaper: yes, definitely, futures. But my conclusion was that UPRO wins by simplicity of leveraging (and granularity) without being overly expensive.
2 Likes
Hi Velli
I don’t quite get what this has to do with leverage. I also don’t know so much about FX-trading. Lastly there also exist government bonds in CHF (of course).
Generally holding cash as cash in an account gives subpar returns as the provider takes its cut. At least consider money-market and short-term government bonds (-fund).
Hi Helix
You are right per se holding cash has nothing to do with leverage. But as I try to look on my overall portfolio asset allocation preferably you have a risky part
e.g. 55% UPRO (3x of SP500) or any pther leverage you prefer (combi UPRO + VOO) and
a negativ correlated part
e.g. 45% TMF (like proposed by HEDGEFUNGIE 3x on US Treaury Bonds but they performed horribly during inflation 
)
Thus i was thinking if it might not make sense to stay with those 45% in CHF cash (ok without leverage unfortunetaly but i don’t know any better )e.g. on wiLLBe at 1.55% interest rate. In a crisis (UPRO decreases significantly) i could imagine that the CHF will get stronger as typically a lot of liqidity goes into CHF in a crisis and then rebalance using the CHF cash to buy UPRO to get back to 55/45%.
What are your thoughts about it? Would be happy and gratful if anybody has a different strategy on this less risky part of the portfolio.
Thx 
If you have 55% Upro and 45% cash you WILL have a bad time during crisis.
The big draw to TMF is that it is also 3x leveraged and tends to go up during crisis/severe market drawdowns. And because it‘s 3x leveraged, you can counteract the severe drawdown (and I cant state this enough, in 2008 for example Upro would have been close to wiped out, meaning drawdown of 90+%).
Yes totally. 10‘000$ invested in „UPRO“ in December 2000 decreased to 574$ in September 2002 (-94.3%). Recovered back to 9‘200$ in October 2007 to get crushed again to 229$ in February 2009 (-97.5%). This doesn‘t include TER or other costs.
Going down from 100k to 2.5k in 1.5 years is as close to a wipeout as you can get.