Shorting Oil, Hedging Tesla

[jhm]

As Tesla shareholders, we are painfully aware of how much the price of oil can move this stock. The stock is correlated with oil in the market whether we whether a fundamental connection exists or not.

Worse yer, oil has become quite volatile since it began falling in mid 2014. It has been in oversupply fir two years and has accumulated an historically high inventory of oil in stirage. The economics of storage tends to place both a lower and upper bound on the price of oil for the foreseeable future. I believe the lower bound is in range of $25 to $30 and the upper bound $40 to $45. These bounds depend on the oil futures curve and the cost and availability of storage capacity so long as the market is in oversupply. Thus, with these caveats, oil is likely range bound.

Tesla investors can use this information to hedge oil sensitivity out of their Tesla position or to make a conscious choice to be net long or net short oil. The purpose of this thread is to discuss oil trade and how to manage oil price risk as a Tesla investor.

 

[jhm]

Short-Term TSLA Price Movements - 2016

Here's a link to my original proposal for hedging TSLA with SCO, a double invest oil ETF.

The basic set up involves regressing multiday returns of TSLA on multiday returns for SCO. The slope of this regression, B, gives you the hedge ratio. If you hold a portfolio of $TSLA - B×$SCO, (in dollar value), this will be uncorrelated to SCO and hence oil.

I have since revisited the question of the number of trading days used for calculating the returns. Initially I selected 25 trading days because I have a 30 calendar day holding rule to follow. But I wanted to see if performance improve with different time horizons. Indeed, a 39 trading day return maximizes back testing over the last 24 months. The annualized return on the adaptively hedged portfolio for 39 days was 24.8%, while for 25 days it was 23.4%. The daily volatility was about 2.1% for both. So I prefer the longer time horizon.

The current beta for 39 day returns is -0.1245. So a hedge portfolio with $10,000 in Tesla would include $1245 of SCO.

This hedging ratio does change a bit over time, so if anyone wishes to try this I would recommend that they build and maintain their own spreadsheet. I'm happy to help someone work this out, but I won't share my spreadsheet. It really is best to develop your own tools so that you know how they work and what assumptions go into them. Best of luck.

 

[SBenson]

James, Thanks for taking the time in posting this. I will spend sometime to see if this something I want to do.

 

[jhm]

Hey, Bensen, I tried to apply the same regression model to SolarCity, and there was virtually no explanatory power.

I also tried SPY and got a 9% Rsq. TSLA has 16% Rsq. So it is rwmarkable that oil has significant impact on both Tesla and the S&P500.

 

[durkie]

Why choose SCO? Those leveraged ETFs always make me nervous with their single-day scope and potential to come out behind even if the underlying is relatively flat.

 

[jhm]

Why choose SCO? Those leveraged ETFs always make me nervous with their single-day scope and potential to come out behind even if the underlying is relatively flat.

What would you recommend?

 

[jhm]

Here's what I like about SCO.

• It's leveraged so I can use less capital to hedge my Tesla position. Naturally you give up a little interest for leverage.
• SCO is easy to trade. It's one of the most subscribed and liquid inverse oil ETFs.
• I do not need a margin account to short oil or need to maintain a reserve which would reduce capital efficiency.
• A double inverse tracks oil to the negative 2 power, allowing for time decay interest. This gives me lots of upside when oil falls but limited downside when oil rises. I do not want the black swan risk of oil going to $299/b for a freakish set of events.
• An ETF is not time bound like an option. So I can hold a hedge as long as I like.
• SCO has strong negative correlation with both Tesla and oil. This makes it an effective hedge instrument.

So for me, SCO is a pretty good hedge instrument. The framework I am suggesting, however, can utilize any hedge instrument. Simply regress Tesla returns on your candidate hedge returns. You can even test out different hedge instruments to see which one gives the best performance. I did try out another oil bear ETF that shorted oil and gas companies rather than oil itself. This instrument had very weak correlation, low Rsq, so I rejected it as a candidate.

An effective hedge instrument need strong negative correlation with both oil and TSLA. If I were only trying to short oil, then the correlation with Tesla would be irrelevant. But as a hedge instrument I need it to efficiently net out whatever sensitivity TSLA may have to oil. The lower the Rsq, the less effective the hedge is. So SCO has a 16% Rsq. If anyone finds a stronger hedge, please let me know. I have not done an exhaustive search, so there may well be stronger hedges.

 

[durkie]

That's exactly what I was wondering -- thanks for spelling it out for me.

 

[jhm]

China sets target for 5 million “new energy” vehicles by 2020

So China wants 5 million EVs on the road by 2020. This is an interesting target from the point of view of how long it will take for EVs to seriously disrupt demand for oil.

Bloomberg suggests that 50 million EVs on the road is the turning point. In my analysis I have focused on 25 million EVs in annual sales as the critical event. As of 2015 the plug-in fleet is 1.3 million. If this continues to grow at 50% per year, we hit Bloomberg's threshold in 2024 and my threshold in 2025. So this is my working scenario. By 2025 oil demand is in structural decline.

So how does China's plan for 5 million EV by 2020 impact my working scenario. Firstly, my working scenario has the global fleet of EVs at 9.9 million in 2020. So if China hits 5 million, it could mean that half of the world's EVs are sold and mostly made in China. Will Germany, the US, Japan, Korea and other auto making countries allow China to get that much of a lead on the EV market? If so, China could be positioned to dominate the global EV market going forward. I don't think that auto making nations or the automakers themselves will want to let China command a 50% market share. So suppose the rest of the world steps up to contain China to 25% market share. Then we are talking about 20 million EVs worldwide by 2020. This accelerates my working scenario by nearly two year, whence oil enters structural decline by 2023.

So China's ambition could accelerate the EV market and the fall of oil. For anyone contemplating shorting oil longterm, this sort of acceleration is worth contemplating. The oil futures curve has oil at $52/b in 2024, but if China accelerates the EV disruption, the actual price in 2024 could easily prove out below $25/b. Thus, under the China scenario, there is a major collapse of the futures curve sometime before 2020. Hard to know when a collapse in expectations could happen, but when it does, the price of oil can fall by 50% or more. Hypothetically, if the Model 3 unveiling were somehow to depress 2024 oil expectations to $25/b, the spot price of oil would fall below $14/b. Now I don't seriously believe that the M3 Unveiling will have that sort of devastating impact. It's only meant as illustration. The only thing proping up the price of oil today is the belief that oil will command a higher price in the future, as this belief motivates traders to buy and store surplus oil. When the futures curve collapses, oil in storage floods into the market depressing prices. The present price floor is all about future expectations, and expectations can change rapidly without warning.

 

[jhm]

OPEC-Russia Meeting Set For April With Or Without Iran | OilPrice.com

Oil is up about 4% today.

Put April 17 on your calendar. OPEC will meet to consider putting a floor under oil. This hope has been fueling the current rally, but this could easily be a buy the rumor sell the news type of play. Likely nothing will come of this meeting and oil prices will fall immediately after.

So I am thinking about the lead up an entry point for shorting oil to hedge Tesla.

 

[jhm]

Oil is edging up to $40. It's at $39.85. So I'm am thinking about starting to accumulate SCO. I think it could get to $42 or higher, so I won't buy in all at once.

Above $40, shale producers will ramp up production that they've been choking or even completing wells. So this combined with huge inventory and continued oversupply leads me to think there is no moderately sustainable price above $40.

 

[jhm]

So I've started to accumulate with SCO at $112.45. A hedge ratio of .125 on TSLA at $225.21 (yeah Tesla!) works out to 1 share of SCO per 4 shares of TSLA. So in my first installment I go just 10%, 1:40. I want to live with this awhile before getting fully hedged.

Wish me luck!

 

[Fallenone]

I think I replied to you in the short term thread when you first proposed this hedging idea. My thought was maybe you can try to short a leveraged positive oil ETF, like UCO. This will also let you capture some of the decay of leveraged securities. Need to calculate if it is worthwhile for the interest fee on shorting though.

 

[jhm]

I think I replied to you in the short term thread when you first proposed this hedging idea. My thought was maybe you can try to short a leveraged positive oil ETF, like UCO. This will also let you capture some of the decay of leveraged securities. Need to calculate if it is worthwhile for the interest fee on shorting though.

I can run a regression model UCO (do you mean USO?) and see how that does. Do you know what the shorting fees would be like?

 

[Fallenone]

I can run a regression model UCO (do you mean USO?) and see how that does. Do you know what the shorting fees would be like?

I meant UCO, 2X leverage of USO. USO is not leveraged. At Fidelity it shows 2% annual interest fee.

 

[BriansTesla]

I really like your idea here. Just bought 100 shares to start.

 

[jhm]

I meant UCO, 2X leverage of USO. USO is not leveraged. At Fidelity it shows 2% annual interest fee.

Ok, thanks!

 

[jhm]

Ok, I tried UCO. Using 39-trading-day returns, I get an Rsq of 24.8% and beta of 0.28. This Rsq compares favorably to the 15.6% for SCO. So this looks promise for someone who can do pairs trading.

The bad news is that my back testing is very unstable. I have to make an assumption of how much cash to hold along shares long TSLA and shares short UCO. In back testing I am looking at daily returns on the portfolio. So I hold a dollar of TSLA, negative beta dollars of UCO and k times beta dollars cash. So using k = 2, the annualized return is -31% (ignoring shorting fees) and the daily standard deviation 2.42%. So there is little reduction in volatility which is 2.68% unhedge, and the return is awful in spite of oil tanking over this period. But what is dreadful is that if I use k = 1.99 instead, the back test annualized return is now positive 66%. So this is enormous sensitivity. Either I am doing something fundamentally wrong here, or the procedure is just not stable. In particular one has unbounded downside risk shorting UCO which means one bad trading day can blow everything up.

So I am not at all comfortable with this. If someone knows how to do this correctly, please let me know. All I know is I can't get this to work for me. With being long SCO there is no need to hold cash and downside risk is finite, so it seems much more robust.

 

[renim]

... it could mean that half of the world's EVs are sold and mostly made in China. ....

It was a file I made to understand 2015 EV charging, but anyway about half the world's EVs are sold and mostly made in China. The momentum is such that in 2016, it should reach about the 50% mark.

as a side note, it also demonstrates that the global EV charging standard is settled, its some Chinese standard that no one uses outside of China but long term, will probably follow China's great railway/shipping push across and down asia and through to europe and africa

 

[jhm]

View attachment 167507

It was a file I made to understand 2015 EV charging, but anyway about half the world's EVs are sold and mostly made in China. The momentum is such that in 2016, it should reach about the 50% mark.

as a side note, it also demonstrates that the global EV charging standard is settled, its some Chinese standard that no one uses outside of China but long term, will probably follow China's great railway/shipping push across and down asia and through to europe and africa

This is helpful. I knew that China was moving on this, but did not realize they were already matching the rest of the world. I hope the West can catch up.