“The log-linear rule falls in between the “traders” price and the “net orders” affecting the price action.” Prof. Doyne Farmer and Prof. Shareen Joshi*
Let’s be absolutely clear: The National Best Bid and Offer (NBBO) is meant to protect the retail trader from pricing manipulation by their counter-party that could cheat them out of their profits, while benefiting the brokerage, proprietary dark pools and/or High Frequency Trading driven by their respective algorithms.
So I ask: Have you ever experience a “price dislocation”? What I mean is that the price that you executed your exit at was blatantly offset so much – say over five cents – that an elephant could walk through it? Well, you’re not alone. Unless you the means of tracking this, say with an excel spreadsheet that is plugged into a brokerage house platform, you won’t know the origin of the “offset” that cut your calculated profits before closing.
EVIDENCE – PRICE CHANGE VALUES
Here is the comparison between the Thinkorswim (TOS) platform’s option chain for AAPL (Apple, Inc) premium prices and my Sharebuilder (SB) close price. I traded AAPL at the APR option chain, the Strike price was 127.14. My entry as shown below from the SB transaction history was a scaled in 2 contracts per 2 trades with a Long Call position.
My target limit was 2.05 on the Ask premium. When 2.05 was hit I executed my close on the SB positions, but was you see, even though there was a drop in the premium on the TOS platform, the SB platform – connected to the BATS exchange – didn’t give me “best price” in accordance to NBBO rules.
Above is the SB transaction history. My closing premium for all 4 contracts was 1.93, not anywhere near what the TOS platform was showing at the same Strike price.
Just below the yellow highlighted ban is the current premium prices, last price and mark price, which all are well above the 1.93 closing price that SB gave me. My profit loss was over $50 on the four contracts, leaving me with $16 in net profit. Moreover, SB’s real-time price at this Strike price was previously held at 1.86 while the TOS platform’s Call premium price increased dramatically toward 2.05. The lag time on the SB to catch up, and then not even matching “best price” as shown on the TOS platform didn’t exceed 1.96.
OFF THE GRID
Whose the culprit behind this? Conniving Leprechauns? Pesky GPU malfunctions? You may be surprised to learn that it boils down to High Frequency Trading algorithms (HFTs) trolling the exchanges for an opportunity to “flash trade” off of your executed trade; either when you open or close. Since HFTs have direct access to contracted co-location to Wall Street exchanges, primarily located in New Jersey, their latency is measured in nanoseconds, whereas yours and mine are measured in milliseconds.
I AM THE WALRAS
If you’re inspired to build an excel spreadsheet that will expose this detrimental price aberration you could turn to Walras’ formula:
dp over dt = -Beta D(p, …)
Or Kyle’s model formula:
P t+1 – P t = ω
But let’s keep it simple, as my point here is to just make you more aware of the fact that virtual pricing is overtaking your “true price” formation. Take into consideration the first element in their formula concerning an asset or option premium’s price formation:
“f” = Form
Form has two components that determine to be path dependent based on the price range (Bid/Offer) inputs provided by any one of the 13 market exchanges at any one second that eventually shows upon your computer screen.
The origin of the asset’s “price” is fundamentally started at the opening of the market session and then sequential accumulation of both Ask and Bid share orders can be calibrated to determine intraday price trends. That means if you set up an excel generator to record any equity Ask and Bid share orders, you can scale out a directional price movement forecast using the appropriate computational statistical mathematics.
According to Farmer & Joshi:
“a. This defines the weighted side which in turn shows the propensity toward formulating a trend that is either Bullish or Bearish.
b. Exponential Distribution equation defines the next calibration that is probability density accumulation.”
High Frequency Traders Can Shift “True Price”
There are 24,000 seconds during the Wall Street scramble to trade in one day session. Within those 24,000 seconds, High Frequency Trading (machines talking to machines) causes 2.4 price dislocations, most often times cheating retail traders out of a few pennies on the equity’s Bid/Offer spread, including option’s premiums. That comes out to 57,600 price dislocations per trading day.
I have built a deterministic trading strategy that incorporates a quantitative excel spreadsheet “real-time” data input, calibrated to a set of parameters that eliminate the “noise” while showing the directional price formation of the underlying asset correlated to the options premium. The most telling function is the price shifts and comparative profit between simulated Call/Put profits during intraday trading that are depicted on the GRTS Tactical Trader.
Below is a screenshot of the GRTS Tactical Trader spreadsheet:**
NBBO CROSS CHECKING
I currently use TD Ameritrade, Thinkorswim, Dough, and Capital One Sharebuilder. The reason is for the very reason I mentioned at the beginning of this article: tracking price dislocations, i.e. “flash trading” by HFTs that cause “virtual price dislocations” giving them a profitable edge on my trade executions, while I take a greater loss.
* “The dynamics of common trading strategies”, Farmer, Doyne, J. & Joshi, Shareen
** For TOS traders, the GRTS Tactical Trader is currently available in the Apache Open Source Excel format, for a token donation. Inquiries via email are welcomed for further information and explanation. This is for advanced users of excel and day traders.