Thursday, July 29, 2010

Switched my short VIX over a month

As the market is selling off today, the premium of Aug VIX over the VIX index has narrowed to under 1 while the difference between Aug and Sept VIX futures remains at 3. So I've covered the short Aug VIX referred to in Uptrending Equity: Volatility arb and opened a new short in Sept VIX. I'm still hedged with puts, and I'm showing a small gain on the overall hedge.

Friday, July 23, 2010


A new inverse VIX ETN is coming out as posted about on VIX and More: XXV and the New VIX ETN Landscape and by Adam Warner at Too Err Is Blog-Human | Blog - Daily Options Report. Outside of roll-yield considerations, I disagree with their assessment that it might be a good long-term hold. This is due to one crucial detail regarding what Adam Warner considers to be the "index" by which the XXV performance is to be calculated. Mr. Warner seems to define the "index" as the performance of VXX, while I assume the "index" will be a more independent measure of VIX performance, without the rebalancing drag and management fee drag. So while Mr. Warner assumes that the relative underperformance of VXX over time will serve to enhance XXV's returns since XXV's performance will be calculated as the inverse of the index's, I believe that VXX's underperformance will not enhance XXV's returns at all.

I'm not saying that a strong contango in VIX futures won't help XXV's returns--it definitely still will. What I'm saying is that I believe XXV and VXX will both fluctuate around an ever-decreasing center of mass, due to rebalancing and fees, that will cause it to be a poor investment choice long term, like many other ETFs/ETNs. Again, to re-iterate, this is aside from roll yield considerations. Basically, over time, if you would like to buy XXV, the better choice will be to short VXX.

See my earlier post Uptrending Equity: Volatility arb to see how I'm profiting from the presently large VIX contango.

Thursday, July 22, 2010


Similar to the time soon after VMW's 2007 IPO when EMC's 87% stake in VMW was worth more than all of EMC's own market cap, VMW has again been climbing in value relative to EMC for the past half year.

My trading instincts tell me that the arb isn't wide-open enough to consider putting on a trade yet, as I prefer more ridiculous blown-out trades to begin reverting before entering, but something to keep an eye on.

I welcome comments regarding EMC's worth outside of its 80% stake in VMW.

BP distressed debt arb

I've been reading about BP's 1-year commercial paper yielding over 10%. I don't have access to a bloomberg or other real-time sources to check out what the prices were a few months ago but that sounds like at least a 50% price drop since pre-oil-spill to me (if someone has a graph of 1-year yields or CDS yields, please post it in comments). If it's paper is down 50%, then logically its equity should be down way more than 50%, right? But it's not, it's only down a little less than 50% since pre-oil-spill. So buy commercial paper (or buy bonds, or short CDSs), short BP equity, and you should make money no matter what happens to BP.

Scenario analysis of two possible extremes:

A) BP is just fine, doesn't pay out hardly anything in environmental suits (pretty much impossible right now). Commerical paper and equity both double back to where they were.
B) BP goes into bankruptcy and can't meet all creditors claims. Equity goes to 0 (actually, for as long as the equity trades, it would likely stay ever so slightly above 0 just b/c stocks simply don't trade down to 0 for as long as they exist, like FNM/FRE right now). Creditors aren't made completely whole, but get some fraction of par, probably a large fraction. Value of commercial paper stays way above 0 b/c a good-sized chunk of it is still paid out.

So as long as you are long fixed income in slightly larger size than you are short equity, you should make out well. However, I would personally tilt the hedge a little more heavily on the short side, as I don't see BP coming back much higher for years at least.

This arb is probably possible because a whole bunch of funds were forced to dump BP's debt at the time it was downgraded, no matter what fair value should be, while equityholders were not forced into such a firesale.

Obscure merger arb

Disclaimer: At the time I last did this analysis, PRS.MC was priced at 1.81 euros, so the evaluations below are made w/ that input. Adjust accordingly for the current market price of PRS.MC.

Short LIA, long PRS.MC, and also get long EUR.USD (=USD/EUR) if you want to hedge currency risk of the trade (as an independent trade, I wouldn't considering going long USD/EUR prudent at this time, as a trendfollowing system, which works over time, would dictate a short position at this time). PRS.MC details here:

Here's the rationale:

LIA is a blank check company. They have a bunch of cash sitting around waiting for the management buyout "pros" to do something with it. They've finally decided to blow it on something. Here's the deal:

On p. 9 of the presentation, you will see that every share of LIA will get 1.173 shares of ordinary PRS.MC, and .563 shares of the NVCS of Prisa.

The 1.173 shares of ordinary Prisa is easy enough to value-> (1.173 shares)*(1.81 Euros/sh)*(1.196 USD/EUR) = around $2.54 USD.

The terms for the NVCS shares are listed on p. 3 of the presentation. It's a little messy, but it looks like for each NVCS, you get a 7% of 7.331 Euros annual dividend for 5 years, and after 5 years, the NVCS will likely be redeemed for ordinary PRS.MC shares at a conversion rate that decreases if more dividends have been paid. If no dividends are paid, the 7.331 "stated value" is retained till the conversion time, and you would basically get almost 2 ordinary PRS.MC shares per NVCS at that time. Essentially, the more dividends are paid, the lower the "stated value" that remains in the NVCS, and the fewer ordinary shares of PRS.MC you'd get when the NVCS is finally converted. Also, Prisa retains the right to choose whether the issue a dividend, thereby decreasing the "stated value," or retaining the "state value" of the NVCS, so Prisa can game the choice to its advantage depending on the price of ordinary PRS.MC shares. This Prisa option decreases the value of an NVCS to the investor from what it would be without the option. With a lower value of PRS.MC shares, Prisa will withhold dividends. In the end, assuming no dividends are issued and the "stated value" of the NVCS remains intact until the end, then you will get almost 2 PRS.MC shares per NVCS. The math they used to calculate the "target value per Liberty share," given their assumptions about the current price of a PRS.MC share and the exchange rate assumption, supports this interpretation.

Problem is, the price of a PRS.MC share has dropped from the assumed price of the presentation, and the EUR/USD exchange rate has dropped as well.

Now you get $2.54 worth of ordinary shares and .563 shares of an NVCS worth 2*.563*1.81 Euros*1.196 USD/EUR = a total of $5.08 of PRS.MC stock for each share of LIA.

Your catalyst for convergence will obviously be when the LIA is actually converted to those other shares, if it doesn't drop in price before then. LIA is currently priced at almost $10. The stock is reasonably thick, so you could probably short at least 100K shares while still being able to get out of all of them fairly easily should something unforseen occur.

I typically dislike merger arb, but in this case, whether the deal goes through or not, the risk of a price increase in LIA is diminished. It was overpriced before it decided to blow its money buying this thing. If you look at the record for blank check companies (SPACs) in general, it's terrible. LIA got on my radar originally for being priced at a steep premium to NAV (basically cash). If the acquisition doesn't proceed, then nothing much should happen to LIA's price.


I've been long BANR for the last couple weeks or so since the stock issuance and massive insider buying at $2, followed by a slight move up out of the 1.90-2.10 area. If you aren't long already, this may be a good spot to pick it up around 2 without risking much. I was hoping for a nice sector short squeeze in regional banks to help my position, as occurred in Jan and Apr of this year, but the mini-sector-squeeze fizzled out pretty quickly this time. However, I'm still following my plan on holding BANR for a big move.

Volatility arb

There is a large contango in VIX futures. I'm arbing it by shorting Aug VIX and buying Sept SPY puts (since the Aug VIX is supposed to reflect the IVs of Sept options) at various strikes w/ IVs at around 22-24%. Over next week or 2, assuming the market doesn't move, Aug VIX should converge down to VIX index, while IVs on SPY options increase as they near expiration. If market moves up or down, the two sides of the pair offset each other. In fact, on the face of it, it appears that the net paired trade is long gamma, meaning that the pair together makes money the further the market moves away from the original price upon entry, assuming both sides of the trade were delta-hedged at the beginning (you will have to come up with an approximation for the delta of Aug VIX). This may lead one to want to sell calls on VIX instead of outright shorting it if you want to neutralize gamma. However, a short VIX position itself may be slightly negative gamma already, as there's a built-in floor to VIX as the market moves up, while VIX can scream much higher as the market falters. Thus, I think an outright short on VIX is adequate.

Alternatively, you short the VXX ETN, which is comprised of both Aug and Sept VIX futures, and buy corresponding weights of Sept and Oct SPY puts. As of July 22, the VXX fund has mostly Aug futures and is dumping some every day to buy Sept futures. In fact, this may be preferable, as people are always trying to front-run these cumbersome ETFs/ETNs.