I am currently writing my thesis about the lack of volatility / changing market structure we see nowadays on the financial market. I believe an important factor in this development is the delta-hedging of the market makers. I saw numerous articles about so-called gamma imbalance from Mr. Kolanovic on ZeroHedge. For example: see http://www.zerohedge.com/news/2017-03-21/kolanovic-today-gamma-imbalance-shifted-toward-puts-first-time-5-months
I tried to reconstruct this graph for the Eurostoxx50 in order to backtest it and get supportive data for my thesis. I was hoping you could give me some feedback about the methodology.
Methodology (using Bloomberg):
- I get the option series for the next upcoming expiration for SX5E options with a moneyness between 0.75 and 1.25.
- Next I get the gamma's of each individual expiration
- I calculate the dollar gamma of each option serie with the following formula:
- Open interest * 100 * gamma * underlying spot (SX5E) ^ 2 / 100.
Next I calculate the gamma imbalance as the difference of each option series between put dollar gamma and call dollar gamma. I do this because the graph in the first link displays "SPX P-C imbalance ($Bn). So I subtract the call dollar gamma from the put gamma dollar.
Next I do this for all months expiring after the upcoming expiration month. Gamma decreases as time to expiration increases, so the two different gamma imbalances fit in the same chart.
Can you give me maybe feedback about this methodology? Is this correct or do you see something incorrect?
More importantly, how can I translate this dollar gamma imbalance to the number of shares that need to be bought / sold as a result of the market-makers option exposure? Is this simply the dollar gamma divided by the price of the underlying? The website (https://squeezemetrics.com/monitor/docs#gex) gives some information about the procedure, but no formula.
Next, I would like to know if my understanding of this whole gamma imbalance is correct
"Following Friday's option expiry, the gamma imbalance shifted towards puts for the first time in about 5 months and the market was 'free' to move again," said Kolanovic.
So, in my understanding Mr. Kolanovic assumes that market makers are long delta and short gamma in this case right? If index goes down, puts become more in-the-money and get a higher delta. Long put option buy insurance and they have little incentive to cover when the market drops. However, short put holders (the market makers) are forced to sell shares to become delta neutral again. So this would increase the magnitude of the initial downwards move. If the gamma imbalance shifts towards call, then dips will be bought and volatility will be reduced as market makers delta-hedge their position by buying low and selling high.
How viable is the assumption that markets-makers are always short gamma? Or how is this actually measurable?