Modified it slightly to illustrate a few things. I picked a more relevant benchmark (SPXL), and graphed your cash to see if all was good. It borrows 20% in August 2011, and tends to throw a few trades in on margin every now and then. If that's intended, great, if it's not intentional it could be something to look at.

I see. Thanks, I haven't noticed it...
Any ideas on hedging against rising interest rates?

there is nothing you can do. sorry. I lost $ over the past 2 days bease TMF fell a lot

Well, I was thinking about the same question these days. I believe that, from now on, with the US interest rates hiking, this strategy would probably be more profitable with some dollar ETF like UUP (or the 3X leveraged UUPT) instead. I mean, dollar has gained some momentum in the past few months, and the tendency, at least economically speaking, would be that it continues to do so in the future, mainly because FED is about to increase the interest rates while most of others developed countries are throwing rivers of money in financial markets in order to stimulate the economy. The other feature of investing in a dollar ETF that is fundamentally important for this strategy to work and minimize its standard deviation is that, it is also considered a "safe heaven" for investments, so, when the outlook of the economy is not so good, the dollar tends to appreciate (the same happens with treasury ETFs such as TMF), while the stock indexes goes down. Buuut, that`s just a guess.

Thanks for the interesting thoughts. In investopedia it points to financials and REIT as sectors that benefit from rising interset rate, but they are positively correlated with SPY so I guess UUPT should be a better try.

As a trader versus a quant/programmer I agree with Vincius. A dollar product, in my opinion, could work quite well. The issue is that you can't really know via backtest due to the previous low interest rate environment. And yes I would avoid financials/REITs as options for the exact reason you mention Han, correlation with $SPY. Either way nice system, interested to see how this develops.

Hi Arthur,

I did a quick simple test using UUPT instead of TMF in the pair trading. But SPXL and UUPT is positively correlated. Both the return and the drawdown are inferior to TMF/SPXL pairing.

Best,
Han

Hey Han, thanks for sharing this. I was wondering how you came up with the dynamic SPXL/TMF ratio based on the ratio of spy200ma/tlt200ma to spy5ma/tlt5ma?

Also, don't know if you have seen this which adds VXX (and its ilk) to manage tail risk:
https://quantstrattrader.wordpress.com/2014/11/03/seeking-volatility-and-leverage/

Thanks for sharing, this article looks great. Using naive strategies published by Harry Long gives significant drawdowns and/or spiky behavior sometimes. However, as mentioned in Harry's article, he has a private strategy with bells and whistles performing much better than the published strategy. My guess is: SPXL/TMF pairing can be a great building block, but perhaps some other recipes are required to reduce beta and limit the drawdown. Well, that's just my guess~~

Hi Rakesh, this rebalancing ratio are tuned by backtest. I try to make the ratio a smooth curve (looks like a 2nd order polynomial) to reduce the chances of fitting the noise.
Clearly overfitting is possible. I don't seriously trust these numbers now.

Best,
Han

Yeah Harry Long's strategies are grossly misinterpreted and marketed in a way to make them seem invincible. That blogger quantstrattrader always does a good job of deciphering through his junk. I will say however that Long's concept can be elaborated on and built into a far more robust strategy but you need to integrate a form of risk management, which I only know how to do on a discretionary basis at the moment, and add more components to the portfolio mix.

As an example, continuing off of our conversation Han, adding defensive leveraged products to the portfolio mix would be good components, a sector such as utilities (UPW / SDP) which should do well in a higher interest rate environment.

Hi Arthur, thanks for the idea. You mean short position in utilities?

Best,
Han

I mean long utilities (although short could work too, just depends on what you're trying to accomplish). In my opinion, and keep in mind again that I'm more of a trader than a programmer or algorithmic trader, you want a portfolio with a set of components that should, in an ideal world, hedge each other relatively well. So by being long utilities, a traditionally defensive sector, you get small upside to very small downside in down markets, and medium upside in up markets. We're talking about the market environment post interest rate hike which, in my opinion, will be a very sideways market. Utilities, as an example, would enable you to potentially smooth out your equity curve by aiding in reducing downside when the markets are moving down and contributing to returns when the markets are moving up (or vice versa if you have a system attempting to profit from downside action). It can work both ways IMO.

Much like the dollar concept from earlier in this thread, however, is that a backtest won't truly tell you how implementing this will perform because the reality is that the market environment since 2010 has been unrealistic and unsustainable. Low interest rates / cheap money doesn't last forever. You mention earlier that UUPT (or dollar in general) is correlated to SPXL (or S&P 500) and that is true as of the last 4-5 years, but that will change with the interest rate decision, much like many other correlations.

A basket portfolio with monthly re balancing that in my opinion could work well for 2015 would be long government fixed income (long TMF or short TMV for structural decline concept), short Russia (variety of products lev and non lev), short Brazil (same as latter), long volatility (no lev), long gold (no lev), long real estate (leveraged) with allocations as follows: 35%, 15%, 15%, 15%, 10%, 10%. Whether you choose leveraged or non leveraged products can be up to the individual but the use of a systematic volatility based risk management (trailing stops for example) is fundamental.

I created the last proposal of Arthur. Off course it wouldn't do well in the past... I'll monitor it for the next while and might turn it on with a small amount

Why do you think the strategy will suffer when rising rates? If you look at EUR/USD the USD has gained momentum in the last three months or so (in my opinion due to traders discounting a rate hike from price). However if we have a look to TLT (long term treasuries), we observe it trades upwards with a very strong momentum since a year ago. My question is, why do you believe TLT will suffer from a rate hike? In my opinion I believe its pretty much the opposite at the moment, in fact some traders are using it as a hadge against rising rates.

PD: In order to hedge against rising rates I would look it from a correlation point of view, but not correlation against an index (such as SP500), but correlation against a factor (rate hike). So from a fundamental point of view I would look to products which rise with rates (postively correlated to factor rates). The best product to use in this case would be Interest Rate Swaps, however, due to volume and collateral restrictions I believe this is option is only available to insititutional investors. Dollar related products are such as UUP are the best option to retail traders. Most institutions are seeing EUR/USD reaching 1.15 or lower in 2015. Also, insitutional research (which unfortunately I cant post here) studying reaction of US equity markets to rate hikes show that equity markets tend to correct when rates effectively rise (not when the announcement is made), but this correction is not a severe one, making the rate hike a reason to buy rather than to sell.