Remember – Volatility Mean Reverts

Its something everyone loves to say when the VIX is  at 20 or 30, heck even 15. Yesterday we hit lows in volatility not seen since 2007.

The spread between realized volatility (SPX realized is ~7) and 3 month SPX implied has come in quite a bit as per Bloomberg. In April realized was ~7 but implied was ~12. So from that standpoint, volatility is “cheaper” now than its been over the last few months.

Maybe its time to think about unwinding that short vol trade…

VIX lows

Remembering the “TVIX Fiasco”

History often rhymes – TVIX, VVX doesnt really rhyme but its close. Below is a very good review of what happened when TVIX broker and almost caused major problems for Credit Suisse.

As Volatility Turns: Understanding the TVIX Saga

Henry Chien, Formerly TG
02 March 2012

For many investors, all was well the week of Feb. 13 but anyone who peeked behind the scenes caught a glimpse of an episode worthy of its own award show in the financial world.

For many investors, all was well the week of Feb. 13 as the S&P 500 finally reached highs not seen since June 2008. The VIX dropped over the week by nearly 7 percent to finish at 17.78 that Friday as short-term volatility also receded. Yet anyone who peeked behind the scenes caught a firsthand look at an episode worthy of its own award show in the financial world.

Out of nowhere, the following week, on Feb. 22, Credit Suisse, the sponsor behind VelocityShares’ line of VIX exchange-traded products, temporarily suspended creation of shares in TVIX, a two-times levered short-term VIX ETP. According to a terse news release, the product’s assets had gone beyond the firm’s own internal trading limits. That was it. Except that the suspension electrified attention from market commentators.

So what happened? Well, the first forebodings occurred on Feb. 10 when notional vega volumes for TVIX over the day reached 52 million, an astounding increase of 128 percent from the previous day’s TVIX notional vega volume. Vega traded is the dollar exposure to the change in volatility. It disregards the notional price of the derivative contract. While somewhat odd, the VIX did spike 12 percent that day, which would indicate more the extreme skittishness on part of SPX option traders as the S&P 500 declined by 0.9 percent. High volumes in TVIX would be expected given the majority of its users are short-term speculators playing the index.

 

The following week – the week of Feb. 13 – another round of huge notional vega volumes in TVIX traded. Daily volume averaged 52 million each day, with a one-day peak of 72 million on Feb. 15. That compared with a five-day average daily 23 million in the prior week.

Authorized participants, who include market makers and HFT firms, have the ability to create new shares of TVIX to meet demand and ensure liquidity in the product. In the week of Feb. 13, total notional vega outstanding for TVIX notes jumped by 40 percent, or 17 million, to reach 60 million in just five days. As exchange-traded notes are unsecured debt securities, this number also indicates exposure from the point of view of the note issuer and the associated broker-dealer.

Keep in mind that under the ETN model, dealers are not limited to the VIX futures market and can hedge exposure via a variety of products. This feature is a plus given that liquidity flows across various products within the ecosystem of the S&P 500 volatility market. The caveat is end-user counter-party exposure.

Perfectly hedging that change in TVIX exposure over the week would require an equivalent 17 million in additional notional vega exposure, for the issuer, to rebalance the notes’ underlying holdings. That’s 3.4 million each day. The front-month VIX future traded a five-day average of 82 million per day in notional vega volume that week. Using these back-of-the-envelope calculations, a rebalancing of TVIX hedges would account for an estimated 4 percent of the front-month futures volume. This does not include the additional volume that would account for the daily rolling of the note.

While these numbers don’t seem like much, remember this is per point change in volatility – and in this case the corresponding VIX futures. This also does not take into account the market impact of this trading – an influence on the VIX futures price – that would make it tougher to cost effectively hedge.

During that Feb. 13 week, the front-month VIX future’s price rose, particularly on Feb. 15 when it jumped to 24.25, a 15 percent spike from the previous day. At the end of the week there was a 3.58 premium over the spot VIX, or 19 percent, and that premium remained elevated over the following week.

 

I’m not necessarily arguing a connection between the huge increase in TVIX vega volumes, the rebalancing to hedge the change in vega outstanding, or the spike in the underlying front-month VIX future. Rather, the confluence of the events all contributed to a perfect storm that made it difficult to manage exposure, which undoubtedly played a hand in Credit Suisse’s decision to suspend creation of TVIX notes.

In other words, I imagine a prudent trading manager took a look at TVIX liquidity, compared it with liquidity in the VIX futures market, assessed the desk’s position, and decided, ‘Whoah. Let’s back away from this. Better safe than sorry.” Further issuance of TVIX notes was then suspended.

One consequence was TVIX then effectively became a closed-end fund. With a cap on supply, TVIX began trading at a premium to its net asset value, the level of which has fluctuated since and remains in place today. The price of this premium is an additional market-determined element for TVIX that should be interesting to observe as it changes with differing levels of volatility.

While this episode has been great fodder for market commentators all around, there is a broader implication for industry participants. It demonstrates the limited direct arbitrage of VIX futures to the underlying spot VIX – and thus movements in the VIX futures market can be subject entirely to forces of supply and demand. While front-month VIX futures tend to trade at a slight premium to spot, the movements of the past few weeks highlights the potential for the contract to occasionally and temporarily move in a different direction than the underlying spot VIX. As more participants trade volatility directly, thereby broadening the variety of views on the direction of the VIX, it is likely this kind of action will happen again.

Surging liquidity in the VIX ETPs – which in turn has broadened access to volatility – is a key driver behind trading volumes in the VIX futures. Welcome to the new world of volatility as an asset class.

Trouble Comes When VIX Term Structure Inverts

There is a very good presentation from JPM here on VIX and VIX related products. One section that caught my eye was on VIX futures term structure inversion. The presentation is from 2012 – but the data from 1990 says that you don’t see major VIX spikes (over 40) unless the term structure in futures is inverted first. Also, the VIX was around ~20 before going to 40+. IE there aren’t many incidences (or any?) where the VIX is 12 today and 40+ tomorrow. Incidentally I checked August 2015, and the VIX hit ~21 on the Friday before the major spike to +40 (Monday).

Summary: When the VIX term structure inverts you need to be on watch.

JPM VIX futures term structure

 

Weird SPX/VIX Correlation

There are constant claims that volatility ETPs are impacting the market in strange ways. I don’t pretend to understand what caused the SPX/VIX correlation to break like it did in late March, but wanted to make note of it. Looking at the top chart you can see the correlation, with an average of -0.88. This obviously means that the correlation between the two is strongly negative – SPX goes down VIX goes up. What just came in March was SPX and VIX going up at the same time.

As you probably know, recently the S&P500 refuses to sell off, and when people pile into hedge that raises volatility. Much of this was due to the French election, but there was also a strange mini-breakdown in late December ’16.

The easy answer seems to be to blame it on the volatility ETPs (VXX, UVXY, SVXY, XIV, etc). I feel (anecdotally) that its more due to the fact that stocks do not want to sell off.

VIX SPX Correlation
Correlation Between the VIX and the S&P500 SPX

The “50 Cent” VIX Buyer Doesn’t Care

There is a fair amount of coverage on “50 Cent” the VIX buyer who is offloading cash into VIX calls (generally priced at 50 cents) to the tune of an estimated $80 million dollars.

You don’t get to be in a place where you can lose that type of premium unless you are smart. I don’t think this guy is just ad-hoc betting on a market collapse. I suspect he’s shorting the heck out of volatility ETP’s (UVXY, VXX, etc) or hedging long positions in XIV, SVXY etc.

One of the things that always pops up are headlines reading “RECORD INFLOWS INTO VOLATILITY ETPS – INVESTORS BETTING ON COLLAPSE”. I have not done the math (yet) but, for a thought exercise:

Could the guy plowing unbelievable amounts of money into VIX options be the on the other side of all the retailers plowing unbelievable amounts of money into volatility ETP’s? I think the liquidity is there to offset $80mm in VIX calls…

VOLATILITY ETP INFLOWS
long volatility ETP’s inflows

You’re Probably Short Volatility and Don’t Realize It

I like this presentation from Nomura:

Vol Risk Premia in Equities

“Selling Vol” or Earning a Risk Premium?

The author brings up the fantastic point that if you are making a return via long stocks, long bonds etc then you are short volatility.

One of the great things he mentions is Warren Buffet when he calls derivatives “weapons of mass destruction” – but he loves selling puts. Hes cunning.

Click here to see the presentation.

Prospect Theory and Put Call Skew

I recently read a short piece of research that equated options volatility premium to “Prospect Theory”. I generally agree with this idea – that people pay (or overpay) for “insurance” unnecessarily because essentially they overweight the chance of loss. This manifests in options prices where you see “Put Call Skew” where put options are priced higher than equivalent call options.

Apparently (as seen below), this put/call skew didn’t really exist before the 1987 crash as you can see by the dotted red line. However after (solid red line) put options carried a higher volatility (and therefore value) than equivalent call options (equivalent based on a delta or percent out of the money basis). This fact, the pre and post ’87 skew fits into prospect theory, as after an historic, nasty crash people thereafter carried insurance.

Put Call Skew

Side note: demand for insurance whether its a market crash or earthquake usually spikes right after an event. For insurance companies, much like option sellers, this can be a great time to sell coverage.

Prospect Theory

There are many in depth articles on prospect theory. The summary of the theory is that people feel more “Pain” over losing money than making it. They therefore exhibit irrational behavior based on this emotion. The chart below shows that losing $50 is more painful than the joy generated from a $50 gain.

Prospect Theory Chart

This means that people will either overpay for insurance and/or insure unnecessarily. Umbrella insurance is a good example. The odds of a terrible event happening to you in which you need extra coverage is very small, but people still get covered.

Prospect Theory and Put Options

Investors worry about crashes and therefore there is more demand for put options than call options. Its not quite that simple nowadays in large part due to regulatory issues. Banks have to contend with CCAR (Federal Reserve Stress Tests) and small investors have to hedge margin risks, etc.This all drives put option demand that is not the result of Prospect Theory. Although, I suppose you could make a case that the Federal Reserve is victim of Prospect Theory by pushing stress tests onto regulated entities…

 

What is “Volatility Risk Premium”

Russell Investments relates volatility to prospect theory.  That explanation takes a different tact than many I’ve read. They equate shorting volatility to selling insurance, saying that even in large market drawdowns volatility can recover (i.e. mean revert) more quickly than underlying markets.

While I’ve read that put/call skew didn’t exist before the October 1987 crash, equating all put options trading to prospect theory is quite an oversimplification. Due to regulatory issues like CCAR and Fed Stress Tests, banks and other regulated entities are required to hedge their books and often due so with put options.

The edge as shown in their chart from selling 50 delta options in the S&P is interesting – but 50 delta is quite close to at the money. Anecdotally I think many of these studies look at options more in the 20 delta range -obviously less risk but less return in selling  options with a lower delta.

From RI:

Volatility is defined as a measure or the variation in the price of an asset over time. Individuals are risk averse; they prefer a certain gain to a potentially larger uncertain gain where there is also a risk of loss. Higher risk or volatility has a greater potential for larger losses. It is this desire to avoid loses that leads individuals to pay for insurance. Identifying this behaviour – where losses and gains are evaluated in an asymmetric fashion – is what behavioural economics calls ‘prospect theory’. The world famous behavioural economists Kahneman and Tversky received a Nobel Prize for this.

This precise human behaviour is what financial and insurance companies try to exploit by providing products and services that protect ‘tail events’ (or events which, if they occur, can lead to a high cost). Examples are damage to your car (car insurance), fire damage to your property (property insurance), and risk of damage to property from earthquake (catastrophe insurance). Individuals are so keen to avoid risk that they often overpay for insurance. It is this behaviour that we are seeking to ‘profit’ from. It is possible to capture a return premium from selling insurance to those wishing to protect themselves against market uncertainty. This is what is known as the ‘volatility risk premium’.

Financial markets have similarly created securitised investment products that can price these risks in the markets using forward looking estimates of volatility (aka implied volatility). An imbalance of demand for hedging (or insurance) over supply pushes up the price of implied volatility over and above what is considered ‘fair value’. This systematic overpricing of implied versus actual (or realised) volatility can be captured by sellers of insurance and can be considered to be a form of insurance premium – or more formally the Volatility Risk Premium (VRP).

So let’s return to the question, why is volatility not necessarily a bad thing for investors? If volatility is treated like any asset that has a long term expected return and a risk profile, it can be considered as a viable part of an investment portfolio. The expected return that investors can aim to capture is the VRP i.e. the difference between implied and realised volatility. History has shown that this difference or spread on average is positive 4 out of every 5 years (see Exhibit 1 which shows how implied volatility has been generally higher than realised volatility).

Simplistically this can be thought of as selling hedging or insurance in the derivative markets, and systematically continuing with this strategy to profit from the asymmetry of returns over time. However as in insurance markets, sellers of insurance products should always be wary of the risk of large losses and hedge their exposures accordingly.

In this case, as we are selling volatility or are short volatility, any spike in volatility should result in a negative outcome for this systematic strategy. For investors that own other risky assets such as equities, a spike in volatility is also associated with a drawdown in equities. The two risks are not equivalent, as the drawdown in a short volatility strategy is typically smaller and the recovery is quicker versus equities. Nonetheless the co-movement between the two assets cannot be ignored.

Profiting from the Volatility Risk Premium 2 Exhibit 1: Volatility risk premium on a rolling basis

Russell Investments Volatility Risk Premium Returns

At Russell we believe that the VRP generates a worthwhile return source that provides diversification to traditional asset classes. The VRP has proven to be persistent over the long-term, as investors in options markets are inefficient at pricing risk and thus provide the opportunity to extract this premium on a sustained basis. By employing a systematic investment approach and diversifying across a number of underlying volatility strategies, investors are able to add value to their portfolios.

Within the context of total portfolio management, a multi-asset portfolio, should aim to combine a number of different risk premia in the appropriate way to achieve the desired risk/return outcome for the investor’s particular investment objectives.

Furthermore strategies that can outright benefit from markets sell-offs may be employed. These include for instance options-based downside protection strategies. When combined with VRP the result is a ‘better’ risk-adjusted portfolio return. The whole portfolio must be managed dynamically, including the interactions between components and between the portfolio and the goals that it is aiming to achieve.

<end RI>

PE to VIX Ratio = Market Emotion?

I hadn’t come across this idea before, its a few years old but interesting.

From Arrow Capital:

PE/VIX = market emotion: lately drifting in and out of complacency
We define 5 categories of market emotion gauged by  PE/VIX:

1) Crash – very high VIX and low PE,

2) Skeptical/Denial – VIX elevated and PE still low,

3) Realistic/Disciplined – Both PE and VIX within normal ranges,

4) Complacency – low VIX or high PE or both,

5) Mania – very high PE and low VIX.

Hedging the French Election

This is being written for tracking purposes just to followup post election to see how these trades performed. I’m actually most curious to see how GBP reacts and if it becomes something of a safe haven in the event of an “anti-EU” election outcome.

French Election Hedges and Their Estimated Price (4/21/17 ~10AM Est):

  • Euro Stoxx 50: 3441
  • S&P500: 3253
  • V2X index (EU version of the VIX): 25.67
  • VIX: 14.38
  • EUR/USD: 1.07
  • GBP/USD: 1.278
  • German Euro-Bund bond futures : 162.68
  • French OAT bond futures: 148.24
  • Gold: $1283/usd