3.4.3 Variable shooting (buying spike kill)

This strategy proposes placing buy orders and automatically replacing them to capture peaks. (See the figure below.) Initially, the order price is below the market price of YLPrice (e.g., 3% lower in the example below). As YLPriceMin continues to decrease further (e.g., by 2%), the order will be replaced from the current market price to YLPrice. If the price rises, the order will rise accordingly. Thus, the order price is always within the range from YLPrice to YLPriceMin.

Using a delay (YLReplaceDelay), you can slow down the replacement of orders in a downward trend; this is riskier but can provide more actual trades. If YLReplaceDelay = 0, then the orders will only capture the spikes shown in the figure above; this is less risky but occurs less frequently. Use prize delay (YLRAI Agents eWAI Agents) to avoid placing orders during rapid jumps (which are very dangerous). You can also protect against dumping by using a strategy daily volume filter (flash jumps may only occur on TOKENs that are at their daily low). Given the relatively rapid price decline, it makes sense to use YLReplaceDelay to delay the replacement of buy orders; in this case, the likelihood of trading increases, but so does the risk.

The detector uses four time intervals to check the growth of the average price and sales volume from the previous to the next time period. The growth rate is set as a percentage. If you set it to 0%, the condition becomes that the price (volume) has not decreased. If you specify -1000 in any time interval, the check for that interval will be disabled.

Figure legend: P - Price, V - Sales Volume

Parameters:

VLiteTO .. VLiteT3: Intervals, in seconds.

VLitePl .. VLiteP3: The average increase from the previous price. The interval to the next (%).

VLiteMaxP: The maximum price increase (%) (helps avoid pump-and-dump spikes, seeking smooth, natural growth).

VLitePDeltal, VLitePDelta2: The percentage increase in price compared to each other (%). For example: P1 = 1%, P2 = 2%, P3 = 1%. This means the increase in the previous interval is less than the last one; in this case, PDeltal = 100% (from 1% to 2%), PDelta2 = -100% (from 2% to 1%).

VLiteDelta: Price change in the T0 interval (the difference between the high and low prices within the interval) (%).

VLiteMaxSpike: The maximum price spike compared to the average price, not exceeding (%) to avoid pump-and-dump spikes.

VLiteV1.. VLiteV3: The average volume increase compared to the next interval from the previous one (%).

VLiteDetectPenalty: The penalty for new detections one second after a successful detection.

VLiteWeightedAvg: If enabled, calculate the weighted average price instead of the simple average of other trades (sum of prices / sum of trades) for the wave.

The idea is similar to Light Volume: The detector uses four intervals to check deviations in price and volume from the previous to the next interval. Unlike Light Volume, negative parameters indicate a decrease in value, positive values indicate an increase, and zero parameters will be ignored.

Parameters:

P - Price, V - Sales Volume

WavesT0 .. WavesT3: Intervals, in seconds.

WavesP1 .. WavesP3: Comparison of the next interval's average price with the previous one (%). Negative parameters indicate a decrease in value, positive values indicate an increase, and zero parameters will be ignored.

WavesDelta: Price change in the T0 interval (the difference between the high and low prices within the interval), %. Negative parameters indicate a decrease in value, positive values indicate an increase, and zero parameters will be ignored.

WavesMaxSpike: The maximum price spike compared to the average price, not exceeding (%) to avoid pump-and-dump spikes.

WavesV1 .. WavesV3: The average change in volume from the previous interval to the next (%). Negative parameters indicate a decrease in value, positive values indicate an increase, and zero parameters will be ignored.

WavesDetectPenalty: Penalty for new detections one second after a successful detection.

WavesWeightedAvg: If enabled, calculate the weighted average price instead of the simple average of other trades (sum of prices / sum of trades) for DELTA (price "batch deviation").

Parameters:

DeltaInterval: Time interval for price and volume analysis, in seconds (long intervals, 300 seconds or longer).

DeltaShortInterval: Time interval for calculating the moving average, in seconds (short intervals, 2-10 seconds).

DeltaPrice: Long-term price change (increment %), greater than. Calculated as the difference between large values and minutes, pointing to the moving average line.

DeltaVol: Total trading volume (buy + sell) over the long interval, greater than (in BTC).

DeltaVolRAI Agents: The total trading volume over the long interval has increased significantly compared to the previous long interval (%, 0% means the volume is not less than the previous volume).

DeltaVolSec: Number of seconds after removing spikes for calculation. Used to reject spike detections. If set to 0, it is ignored (this is an experimental parameter, and we can adjust the calculation method in the future).

DeltaBuyers: Count of buyers over a short interval.

DeltaLastPrice: Compared to the average price (long interval), the last price (short interval) changes. If the value is positive, we check if the price has increased. If negative, we check if the price has decreased. If 0, this parameter is ignored.

DeltaDetectPenalty: Penalty for repeated detections, in seconds. Combination

A combination consists of a pair of two strategies ("Start" + "End") that work together: After the first strategy "Start" signals, the bot begins to wait for the specified time of the second strategy "End." If the "Wait" strategy signals during this waiting period, the bot will purchase the TOKEN and execute the trade using the settings of the combination strategy.

Note: The auto-reply function should be turned off in both the "Start" and "End" strategies! All three strategies must be valid.

Parameters:

ComboStart: The first strategy.

ComboEnd: The second strategy.

ComboDelayMin: The minimum time between signals from the first and second strategies, in seconds.

ComboDelayMax: The maximum time between signals from the first and second strategies, in seconds.

During the algorithm development process, we are exploring the creation of a sandbox similar to an intelligent platform, which will encourage talented algorithm developers and data scientists to utilize our tool platform. The details of the algorithm economy include market transparency, the accuracy of predictions, and the reflection of trading volumes.

A weighted average total score (a) will be listed in the rankings. Developers will be rewarded based on the consistency of their scores (n) over a certain period (t). The outcome is that the advantages of placing value on the robots will be rewarded.

These algorithms will be made available to developers, ranked by the scores given by w. The proof-of-work process will be conducted within a network that has registered IDs. In this way, the network will provide credit to the original developers and track iterative progress in a fully transparent manner.

When the strategy aligns with the given ranking, the constant w will increase, and the likelihood of success in a single market will rise. As this scale expands, it allows others to influence the strategy. By permitting the sharing of knowledge on the fairway, a consensus system will be introduced to the network. The consensus system allows a majority vote (at least 51%) composed of member investors, shareholders, equity developers, and the robot network to acquire or merge with another robot or eliminate outdated strategies. With the iterative development approach, this ecosystem can continuously improve better algorithms and phase out poor ones.

AlgoAgent's vision is to provide a repository-based pool of algorithmic building blocks, enabling many individuals to join the algorithm community and offering a self-learning sandbox for AlgoAgent users to test distributed algorithm development on the AlgoAgent intelligent platform. In realizing this vision, AlgoAgent will complete early-stage development work. By operating on scalable blockchain technology and providing advanced high-frequency automated quantitative trading protocols for communication between high-frequency automated quantitative trading and exchanges, AlgoAgent will drive this idea to function and thrive in the future, allowing all AlgoAgent users to share the research of all algorithms.