The majority of financial market trading is performed by machines. These computer algorithms are designed to arbitrage and hedge the markets along many different timeframes.
As market makers and large traders create positions, they will often arbitrage and hedge with delta-neutral portfolios. These portfolios are instantaneously and dynamically adjusted in micro-seconds.
As option expiration approaches, the computer algorithms must unwind and roll forward their delta hedges. This creates order flow which tends to create compression between market delta and market price.
Our historical data shows that the price of a stock index or commodity will often revert to the point of delta-neutral. The can be seen above as the convergence between the red and green lines prior to the option expiration date.
The algorithms which actively hedge delta must also monitor their gamma exposure. When market gamma spikes, market makers may be forced to buy or sell, such as in the natural gas market (Oct 2018) and in the S&P index (Dec 2018).
In November 2018, Bloomberg published an article which described how "negative gamma" was a catalyst in the large decline in crude oil prices in Q4. While Bloomberg highlights the importance of this information, it cannot be found on the $25,000+ per year Bloomberg terminal.