The most dominant market making programs we see today tend to be arbitrage based. They maintain a good spot in the queue, and know when they should take fills. This usually involves a correlation or value calculation. When the market trends it is often because the market maker no longer wants to take fills on one side of the market because their value calculation is moving where they're willing to take fills. After accounting for the probability of informed trading within the flow such methods should yield an average profit equal to the spread for each contract they trade.

So to identify these kinds of moves before they start I think it would be useful to develop a way to detect correlated order flow. Instances where significant market orders and liquidity changes in two correlated assets match up very closely.

SINCE IT WAS SO MUCH "FUN" DESIGNING A MINI INVENTORY RISK ANALYZER...

Perhaps we could do some speculating and theorizing; and come up with a design
for an Indicator to tell us when Market Maker is "Pushing" the market either
UP or DOWN.

Such an idea could relate to the start of a Trend, or the imminent turning of
a Trend which could be more precise than knowing Inventory and Risk levels.

HOWEVER, THERE ARE MANY ISSUES WITH THIS; not the least of which will
be whether you have a QUALITY Market Data feed or not.

I use the Rithmic Market Data Feed which is "unaggregated" and has a
"very high update rate". Most of you do not have such a feed. Just sayin'...

However, theoretically we might still be able to "coax" out something of an
Edge, even though most feeds are not necessarily the highest quality. After
all, there is no feed fast enough to keep up with what is happening inside
the Exchange's Matching Engine; so every feed is necessarily incomplete
to varying degrees.

SO I NEED SOME REAL ENTHUSIASM FROM YOU GUYZ BEFORE I GO ANY FURTHER
with this (difficult) exercise.

FAILING ENTHUSIASM, I'll just stick with Theory and Rationale for a Design, but
probably won't go to the trouble of implementing anything...

Our little Trade Flow Risk Indicator used only Time and Sales, for which
most of you have an adequate feed. This one, at least in my Custom
Implementation, uses the Rithmic Market Depth feed; which delivers up to 1000
updates per second, and that's way out of scope for this exercise.

ASSUMPTIONS:
1) Market Maker moves the Market price (forcibly), and over-powers any
"Retail Pressure" to the contrary, through the sheer power which MM enjoys.

Right off the bat, there will be those who don't or won't accept this simplifying
assumption; but I'd urge you to provisionally accept it, to see where it leads us.

2) A "Wave of Bid and Offer increased Volume" PUSHES the Market Price either
upwards or downwards, and it is THIS WAVE which we'd like to "estimate"
or meaningfully detect.

CONSIDER A RISING PRICE:

Market Maker wishes to lift the Price, so what does She do ? Let's keep it simple
as a concept. In fact MM will need to do 2 things:

a) Add more Volume to the current Best Bid or, more significantly, Raise the
current Best Bid Price by INSERTING new volume in a BID to BUY, in case Retail
Players are willing to SELL (which some players will) and MM intends to BUY
from ANY Retail Sellers, no matter who it is. On BUYing from any Retail Sellers,
MM has just acquired INVENTORY, which she will want to resell at a HIGHER
Market Price. BUT WAIT? How does she get a Higher Market Price? Obviously,
she just plows HIGHER in Price with more BID Volume. So she will Over-Power
any Retail Pressure to the contrary; and FORCE or "push" the Market Bid Price higher.

(Hey, wait a minute !!! Has MM just INCREASED the Value of what she just
Bought at a lower price ??? THAT JUST AIN'T FAIR. But this is exactly what
happens. MM is Forcibly lifting the Market, increasing the value of anything
She's just bought from any Retail Sellers, thus taking No Risk.)

b) Now, what must MM also be doing on the ASK/Offer side of the Market? Obviously
Reducing or Removing her Volume at the ASK, leaving only any Retail Stragglers
who might be on the Best ASK; so that MM Abandons her ASK pressure, and
"lifts" the ASK, transferring her larger Volumes higher in Price, and away from
the Rising Market.

SO THIS IS WHAT WE HYPOTHESIZE IS HAPPENING HERE. This gives rise to
some limited predictions. If MM "wants" to move the Market to a Higher Price,
She will "up the ante" by Raising the Best BID and, crucially, using a STRONG
VOLUME as MM INSERTS her higher priced BID Volume. And on the top side,
which is "Retreating", MM is removing Volume at or near the ASK, and transferring
her Volume to Higher Price Levels. Basically "getting out of the way of her
up-move" by removing ASK tiers to higher Prices, and lower quotes Volumes.

I'll let y'all have a Think about what I'm hypothesizing here; and also if you can
"indicate" (pun intended) your interest in an Indicator that MIGHT estimate
or measure such a process...

Don't be a Hater; but I'm sure there are many who either disagree with this,
or perhaps never thought about it that much... Awaiting your comments.

hyperscalper