market-microstructure-traditional

Name: market-microstructure-traditional
Author: agiprolabs

203installs53forks

How to install market-microstructure-traditional?

agiprolabs/claude-trading-skills/market-microstructure-traditional

$npx -y skills add agiprolabs/claude-trading-skills --skill market-microstructure-traditional

Installs into the current project.

Summary

Traditional market microstructure concepts applied to crypto — order book dynamics, market making theory, price formation models, execution quality measurement, and CEX vs DEX structural differences

Files · 1

View on GitHub

SKILL.md

1	# Market Microstructure (Traditional)
2
3	Market microstructure studies how orders become trades and how trades become
4	prices. Understanding these mechanics is essential for execution optimization,
5	market making, and detecting informed flow. This skill covers limit order book
6	(LOB) theory as applied to crypto markets on centralized exchanges, and compares
7	LOB mechanics to the AMM-based structure of DEXes.
8
9	## Core Concepts
10
11	\| Concept \| What It Tells You \|
12	\|---\|---\|
13	\| Bid-ask spread \| Cost of immediacy — how much you pay to trade now vs later \|
14	\| Price impact \| How your order moves the market price \|
15	\| Order book imbalance \| Short-term directional predictor from queue sizes \|
16	\| Adverse selection \| Risk of trading against informed counterparties \|
17	\| Inventory risk \| Market maker exposure from accumulated positions \|
18	\| Execution quality \| How well your fills compare to a benchmark \|
19
20	---
21
22	## Bid-Ask Spread Decomposition
23
24	The bid-ask spread is not a single thing. It decomposes into three components
25	(Roll, 1984; Glosten & Harris, 1988):
26
27	1. Adverse selection — compensation for trading against informed traders
28	2. Inventory holding — compensation for carrying risk
29	3. Order processing — fixed costs of providing liquidity (fees, infrastructure)
30
31	### Spread Measures
32
33	```python
34	# Quoted spread: what you see on the order book
35	quoted_spread = best_ask - best_bid
36	quoted_spread_bps = (best_ask - best_bid) / midprice * 10_000
37
38	# Effective spread: what you actually pay (accounts for price improvement)
39	effective_half_spread = abs(trade_price - midprice_at_trade)
40	effective_spread_bps = effective_half_spread / midprice_at_trade * 10_000
41
42	# Realized spread: market maker's actual profit (after price moves)
43	# Measured at trade_price vs midprice N seconds later
44	realized_spread = trade_sign * (trade_price - midprice_after_delay)
45	```
46
47	The effective spread matters most for execution quality. The difference
48	between effective and realized spread measures adverse selection — what the
49	market maker loses to informed flow.
50
51	---
52
53	## Price Formation Models
54
55	### Glosten-Milgrom (1985)
56
57	A sequential trade model where the market maker sets bid and ask prices to
58	break even against a mix of informed and uninformed traders.
59
60	- Market maker quotes reflect expected value conditional on trade direction
61	- Spread exists purely due to adverse selection
62	- Prices converge to true value as information is revealed through trades
63
64	Key insight: the spread is wider when:
65	- Probability of informed trading (PIN) is higher
66	- Information asymmetry is larger
67	- Uninformed trading volume is lower
68
69	### Kyle's Lambda (1985)
70
71	Kyle models a single informed trader, noise traders, and a market maker.
72	The market maker sets price as a linear function of net order flow:
73
74	```
75	price_change = lambda * net_order_flow
76	```
77
78	Lambda (λ) measures permanent price impact per unit of signed volume.
79	Higher lambda = less liquid market. Lambda is estimated by regressing
80	price changes on signed volume:
81
82	```python
83	import numpy as np
84	from numpy.linalg import lstsq
85
86	def estimate_kyle_lambda(
87	price_changes: np.ndarray,
88	signed_volumes: np.ndarray,
89	) -> float:
90	"""Estimate Kyle's lambda from trade data.
91
92	Args:
93	price_changes: Midprice changes between trades.
94	signed_volumes: Trade volume * trade_sign (+1 buy, -1 sell).
95
96	Returns:
97	Estimated lambda (price impact per unit volume).
98	"""
99	X = signed_volumes.reshape(-1, 1)
100	beta, _, _, _ = lstsq(X, price_changes, rcond=None)
101	return float(beta[0])
102	```
103
104	See `references/price_formation.md` for full model derivations and the PIN
105	model for measuring informed trading probability.
106
107	---
108
109	## Price Impact Models
110
111	### Temporary vs Permanent Impact (Almgren-Chriss)
112
113	When executing a large order:
114
115	- Temporary impact — price displacement that reverts after your order.
116	Caused by consuming standing liquidity.
117	- Permanent impact — information content of your trade that moves the
118	equilibrium price. Does not revert.
119
120	```
121	total_impact = permanent_impact + temporary_impact
122	permanent = gamma * (shares / ADV)
123	temporary = eta * (shares / time_horizon) ^ alpha
124	```
125
126	Typical alpha values: 0.5-0.7 (square root impact is a robust empirical finding).
127
128	### Square Root Impact Law
129
130	Empirically, price impact scales as the square root of order size relative
131	to daily volume:
132
133	```python
134	def square_root_impact(
135	order_size: float,
136	daily_volume: float,
137	volatility: float,
138	impact_coefficient: float = 0.1,
139	) -> float:
140	"""Estimate price impact using the square root model.
141
142	Args:
143	order_size: Number of units to trade.
144	daily_volume: Average daily volume.
145	volatility: Daily return volat

People who liked this also viewed

6 picks

Type

market-microstructure-traditional

203installs53forks

View on GitHub

How to install market-microstructure-traditional?

agiprolabs/claude-trading-skills/market-microstructure-traditional

$npx -y skills add agiprolabs/claude-trading-skills --skill market-microstructure-traditional

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Market Microstructure (Traditional)
2
3	Market microstructure studies how orders become trades and how trades become
4	prices. Understanding these mechanics is essential for execution optimization,
5	market making, and detecting informed flow. This skill covers limit order book
6	(LOB) theory as applied to crypto markets on centralized exchanges, and compares
7	LOB mechanics to the AMM-based structure of DEXes.
8
9	## Core Concepts
10
11	\| Concept \| What It Tells You \|
12	\|---\|---\|
13	\| Bid-ask spread \| Cost of immediacy — how much you pay to trade now vs later \|
14	\| Price impact \| How your order moves the market price \|
15	\| Order book imbalance \| Short-term directional predictor from queue sizes \|
16	\| Adverse selection \| Risk of trading against informed counterparties \|
17	\| Inventory risk \| Market maker exposure from accumulated positions \|
18	\| Execution quality \| How well your fills compare to a benchmark \|
19
20	---
21
22	## Bid-Ask Spread Decomposition
23
24	The bid-ask spread is not a single thing. It decomposes into three components
25	(Roll, 1984; Glosten & Harris, 1988):
26
27	1. Adverse selection — compensation for trading against informed traders
28	2. Inventory holding — compensation for carrying risk
29	3. Order processing — fixed costs of providing liquidity (fees, infrastructure)
30
31	### Spread Measures
32
33	```python
34	# Quoted spread: what you see on the order book
35	quoted_spread = best_ask - best_bid
36	quoted_spread_bps = (best_ask - best_bid) / midprice * 10_000
37
38	# Effective spread: what you actually pay (accounts for price improvement)
39	effective_half_spread = abs(trade_price - midprice_at_trade)
40	effective_spread_bps = effective_half_spread / midprice_at_trade * 10_000
41
42	# Realized spread: market maker's actual profit (after price moves)
43	# Measured at trade_price vs midprice N seconds later
44	realized_spread = trade_sign * (trade_price - midprice_after_delay)
45	```
46
47	The effective spread matters most for execution quality. The difference
48	between effective and realized spread measures adverse selection — what the
49	market maker loses to informed flow.
50
51	---
52
53	## Price Formation Models
54
55	### Glosten-Milgrom (1985)
56
57	A sequential trade model where the market maker sets bid and ask prices to
58	break even against a mix of informed and uninformed traders.
59
60	- Market maker quotes reflect expected value conditional on trade direction
61	- Spread exists purely due to adverse selection
62	- Prices converge to true value as information is revealed through trades
63
64	Key insight: the spread is wider when:
65	- Probability of informed trading (PIN) is higher
66	- Information asymmetry is larger
67	- Uninformed trading volume is lower
68
69	### Kyle's Lambda (1985)
70
71	Kyle models a single informed trader, noise traders, and a market maker.
72	The market maker sets price as a linear function of net order flow:
73
74	```
75	price_change = lambda * net_order_flow
76	```
77
78	Lambda (λ) measures permanent price impact per unit of signed volume.
79	Higher lambda = less liquid market. Lambda is estimated by regressing
80	price changes on signed volume:
81
82	```python
83	import numpy as np
84	from numpy.linalg import lstsq
85
86	def estimate_kyle_lambda(
87	price_changes: np.ndarray,
88	signed_volumes: np.ndarray,
89	) -> float:
90	"""Estimate Kyle's lambda from trade data.
91
92	Args:
93	price_changes: Midprice changes between trades.
94	signed_volumes: Trade volume * trade_sign (+1 buy, -1 sell).
95
96	Returns:
97	Estimated lambda (price impact per unit volume).
98	"""
99	X = signed_volumes.reshape(-1, 1)
100	beta, _, _, _ = lstsq(X, price_changes, rcond=None)
101	return float(beta[0])
102	```
103
104	See `references/price_formation.md` for full model derivations and the PIN
105	model for measuring informed trading probability.
106
107	---
108
109	## Price Impact Models
110
111	### Temporary vs Permanent Impact (Almgren-Chriss)
112
113	When executing a large order:
114
115	- Temporary impact — price displacement that reverts after your order.
116	Caused by consuming standing liquidity.
117	- Permanent impact — information content of your trade that moves the
118	equilibrium price. Does not revert.
119
120	```
121	total_impact = permanent_impact + temporary_impact
122	permanent = gamma * (shares / ADV)
123	temporary = eta * (shares / time_horizon) ^ alpha
124	```
125
126	Typical alpha values: 0.5-0.7 (square root impact is a robust empirical finding).
127
128	### Square Root Impact Law
129
130	Empirically, price impact scales as the square root of order size relative
131	to daily volume:
132
133	```python
134	def square_root_impact(
135	order_size: float,
136	daily_volume: float,
137	volatility: float,
138	impact_coefficient: float = 0.1,
139	) -> float:
140	"""Estimate price impact using the square root model.
141
142	Args:
143	order_size: Number of units to trade.
144	daily_volume: Average daily volume.
145	volatility: Daily return volat

People who liked this also viewed

6 picks

Type