Scribd
Upload
200 views13 pages

What Is The Microprice?: Sasha Stoikov

The document defines the microprice as the expected midprice in the distant future based on order book information. It presents a Markov model approach to estimate the microprice as a function of quote imbalances. Empirical analysis of stock data finds the microprice converges over time and has flatter volatility patterns than alternatives like the midprice or naive microprice.

Uploaded by

unknown
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
200 views13 pages

What Is The Microprice?: Sasha Stoikov

The document defines the microprice as the expected midprice in the distant future based on order book information. It presents a Markov model approach to estimate the microprice as a function of quote imbalances. Empirical analysis of stock data finds the microprice converges over time and has flatter volatility patterns than alternatives like the midprice or naive microprice.

Uploaded by

unknown
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Motivation Definition Estimation Conclusion

What is the microprice?

Sasha Stoikov

Cornell University

Bloomberg Lightning Talk


Motivation Definition Estimation Conclusion

HFTs care about quote imbalances


qtb
Imbalance It = qtb +qta
, where qtb and qta are the bid and ask sizes
Motivation Definition Estimation Conclusion

Why HFTs care about quote imbalances

This strength of this signal will not go away!


Motivation Definition Estimation Conclusion

What people call the microprice

b a
• Mt = pt +p
2
t
is the midprice
a b
• st = (pt − pt ) is the spread
b
• It = bqt a is the imbalance
qt +qt
• The Gatheral and Oomen (2009) microprice:

ptnaive−micro = Mt + st (It − 0.5)

• There is no guarantee that this price is ”fair” or ”efficient”


Motivation Definition Estimation Conclusion

Approximating prices

Define a sequence of martingale prices

pti = E [Mτi |Ft ]

where
• Ft is the information contained in the order book at time t.
• τ1 , ..., τn are (random) times when the midprice Mt changes
• Assume the 2 dimensional process (Mt , It ) is Markov
• The distribution of Ms − Mt is independent of the level Mt
Motivation Definition Estimation Conclusion

First approximation

pt1 := E [Mτ1 |Ft ]


= Mt + E [Mτ1 − Mt |Mt , It ]
= Mt + g 1 (It )

with

g 1 (x) := E [Mτ1 − Mt |Mt , It = x]


= E [Mτ1 − Mt |It = x]
Motivation Definition Estimation Conclusion

Second approximation

pt2 := E [Mτ2 |Ft ]


 
= E E [Mτ2 |Fτ1 ] Ft
= E pτ11 |Ft
 

= E g 1 (Iτ1 ) + Mτ1 |Ft


 

= pt1 + E g 1 (Iτ1 )|Ft


 

= Mt + g 1 (It ) + g 2 (It )

with

g 2 (x) := E g 1 (Iτ1 )|It = x


 
Motivation Definition Estimation Conclusion

Micro price definition

The i-th approximation


i
X
pti = Mt + g k (It ) = Mt + G i (It )
k=1

where
g i+1 (x) = E g i (Iτ1 )|It = x , ∀j ≥ 0
 

Definition
The MICROPRICE
ptmicro = lim ptn
n→∞
Motivation Definition Estimation Conclusion

Estimation

Further assumptions:
• It takes discrete values 1 ≤ i ≤ 5
• The spread is fixed at 1
• Mt+1 − Mt takes values in (−1, 0, 1)
Goals:
1 Estimate the transition probability matrix
2 Estimate the sequence of functions G i (x) where

Pti = Mt + G i (It )
Motivation Definition Estimation Conclusion

Convergence of G i for BAC


Motivation Definition Estimation Conclusion

Microprice time series


Motivation Definition Estimation Conclusion

Volatility signature plot BAC


Motivation Definition Estimation Conclusion

Summary

1 Have defined the microprice as the expected midprice in the


distant future
2 When fitting a Markov model, we have conditions that
ensures this microprice converges
3 Microprice seems to have flatter volatility signature plots than
midprice or naive microprice

You might also like