VolumeEdge: Documentation

Technical Indicators Simulation Algorithm

The Institutional Execution Simulation Algorithm is a proprietary model used to estimate the real-world market impact and slippage costs of executing large block trades.

This is a VWAP (Volume Weighted Average Price) Execution Algorithm. As an industry standard, approximately 70% of large institutional funds utilize VWAP-based algorithms to pace their orders, ensuring they blend into the market without causing excessive price spikes or signaling their intentions.

By blending the industry-standard Almgren-Chriss framework with our hyper-accurate Volume Profile, we can simulate how a large order would have been filled historically minute-by-minute.

System Hard Constant Settings (Universal Constant)

Minimum price tick size: Tick_Size = 0.01
Ignition impact index: β = 0.5 (compliant with the academic and industry-recognized square root impact rule [1.2.5])
Market Shock Sensitivity Constant: Y = 0.6

The 6-Phase Simulation Engine

1Phase One: Data Preprocessing and Historical Profiling

Before the simulation begins, the following baselines are calculated using rolling historical data (past 20 days):

1. Historical Average Daily Trading Volume (V_ADV): The stock's historical average daily total trading volume.
2. Daily Volatility (σ_daily): The standard deviation of historical daily logarithmic returns.
3. Historical Weights of Intraday Trading Volume for the Previous 10 Days (W̄ = {w̄₁, ..., w̄_T}):
An array of length T (number of trading minutes), representing the historical average percentage of trading volume for each minute relative to the total daily volume.
Cumulative Historical Weights: W_t = Σ w̄_i
(representing the percentage of the total daily trading volume that has been completed historically by minute t)

2Phase Two: Inter-Day Scheduler

Before the market opens each day d, determine today's target participation rate and initial predicted trading volume based on the remaining task volume and remaining time:

1. Q_{rem, d}: Total number of shares remaining unpurchased. If it is 0, the simulation ends.
2. Calculate the long-term target participation rate (α_d):
The participation rate is input by users. This represents the "cruising participation rate" that we should maintain on average each day starting today.
3. Maximum Participation Rate per Minute (α_max):
The participation rate is input by users, representing the highest participation rate we expect in each minute.
4. Initialize Today's Estimated Trading Volume:
V_d^pre_trade = V_ADV

3Phase Three: Intraday Volume Predictor

At minute t during the trading session on day d, we update today's expected total trading volume V̂_d(t) based on the trading volume already occurred:

1. Calculate Today's Volume Multiplier F_t:
F_t = V^actual_{≤ t} / (W_t · V_d^pre_trade)
2. Use Bayesian Contraction Shrinkage Multiplier:
F̂_t = W_t · F_t + (1 - W_t) · 1.0
3. Obtain the real-time revised estimated trading volume for the entire day V̂_d(t) at minute t:
V̂_d(t) = V_d^pre_trade · F̂_t

4Phase Four: Intra-Day Scheduler & Pacing

At minute t, we combine the "Pacing Participation Rate α_d" and the "Estimated Trading Volume V̂_d(t)" to calculate how many shares q_t to buy in this minute:

1. Update today's dynamic total target volume Q_d(t): (Buy more on high volume, buy less on low volume)
Q_d(t) = α_d · V̂_d(t)
2. Calculate the progress deficit rate (e_t):
e_t = [ (Q_d(t) · W_t) - q^filled_{< t} ] / Q_d(t)
e_t > 0 represents lagging behind; e_t ≤ 0 represents ahead of schedule.
3. Calculate the base allocation for this minute (throttle q^base_t):
q^base_t = ( Q_d(t) - q^filled_{< t} ) · [ w̄_t / (1 - W_t-1) ]
4. Calculate the safe speed limit (brake Brake_t): (using the maximum participation rate per minute α_max)
Brake_t = α_max · V̂_d(t) · w̄_t
5. Obtain the safe target purchase amount for this minute q^safe_t:
q^safe_t = min( q^base_t , Brake_t )

5Phase Five: Adaptive Splitter

The algorithm dynamically splits q^safe_t into appropriate percentages of limit orders based on current progress and time urgency:

1. Calculate the current penalty sensitivity K_t:
K_t = K₀ · ( 1 + γ · t / T )
Recommended values: Initial sensitivity K₀ = 4.0, time acceleration factor γ = 2.0.
2. Calculate the limit order ratio x_t for this minute: (preset base ratio x_base = 0.8)
x_t = max( 0.0, min( 1.0, x_base - K_t · e_t ) )
3. Split the order targets for limit and market orders:
Limit order target: q^pass_t = q^safe_t · x_t
Market order target: q^act_t = q^safe_t · (1 - x_t)

6Phase Six: Pricing & Execution Simulation Engine

Enter the physical transaction simulation and calculate the actual average transaction price and volume for this minute:

1. Calculate the sensitivity η and midpoint price P^mid_t for this minute:
η = 0.6 · σ_daily
P^mid_t = (H_t + L_t + C_t) / 3
2. Calculate the dynamic spread and bid/ask prices (rounded to the nearest 0.01 yuan):
Spread: S_t = max( 0.01, 0.05 · (H_t - L_t) )
Bid price (limit): P^bid_t = round( P^mid_t - S_t / 2 )
Ask price (market): P^ask_t = round( P^mid_t + S_t / 2 )
3. Simulated limit order execution determination: (read the lowest price L_t of current K-line)
If L_t < P^bid_t → filled_pass = q^pass_t
If L_t > P^bid_t → filled_pass = 0
If L_t == P^bid_t → filled_pass = q^pass_t · Uniform(0.2, 0.5)
Unfilled Residual Volume: q^unfilled_t = q^pass_t - filled_pass
4. Simulated Market Order and Supplementary Order Execution:
Total Market Orders Executed q^total_active_t = q^act_t + q^unfilled_t
Calculate the Market Execution Price with Slippage:
P^{active_execution}_t = P^ask_t × [ 1 + η · √( q^total_active_t / V_t ) ]
5. Synthesize the Final Score for This Minute:
The actual total buy volume in this minute (100% completion rate):
q^filled_t = filled_pass + q^total_active_t
The actual average buy price in this minute:
P_{execution, t} = [ (filled_pass · P^bid_t) + (q^total_active_t · P^{active_execution}_t) ] / q^filled_t
The true market closing price after being pushed up by your trade:
(simulated footprint price)
P̃_t = C_t × [ 1 + η · √( q^total_active_t / V_t ) ]

What the Simulation Output Shows You

Once the simulation completes, the algorithm returns a comprehensive minute-by-minute execution log and a set of final cost metrics. This allows you to evaluate exactly how expensive it would be to enter or exit a massive position.

1. Institutional Holding CostThe final weighted average execution price of your entire simulated order. This price is drawn as a dotted purple line across the chart, serving as your absolute cost basis. If the current market price is significantly far from this line, you can immediately identify how much paper profit or loss the institution is carrying.
2. Average Market ImpactExpressed in percentage terms (%), this shows how much you artificially pushed the market price against yourself due to slippage and aggressive order routing. A high market impact means the stock was too illiquid for your order size.
3. Minute Execution TableA complete footprint of the algorithm's decisions:
- Execution Volume: The exact number of shares filled in that minute.
- Market Px: The original market close price before your trade impacted it.
- Average Px: The actual blended fill price you received (combining limit and market orders + slippage).
- PART. %: The participation rate for that specific minute (your execution volume divided by total market volume).