Simplified Avellaneda-Stoikov Market Making by Crypto Chassis Open Crypto Trading Initiative

At this point the trained neural network model had 10,000 rows of experiences and was ready to be tested out-of-sample against the baseline AS models. Following the approach in López de Prado , where random forests are applied to an automatic classification task, we performed a selection from among our market features , based on a random forest classifier. We did not include the 10 private features in the feature selection process, as we want our algorithms always to take these agent-related (as opposed to environment-related) values into account.

Bid and ask sizes at the top of the order book provide information on short-term price moves. Drawing from classical descriptions of the order book in terms of queues and order-arrival rates (Smith et al ), we consider a diffusion model for the evolution of the best bid/ask queues. We compute the probability that the next price move is upward, conditional on the best bid/ask sizes, the hidden liquidity of the market and the correlation between changes in the bid/ask sizes. The model can be useful, among other things, to rank trading venues in terms of the “information content” of their quotes and to estimate the hidden liquidity in a market based on high-frequency data. We illustrate the approach with an empirical study of a few liquid stocks using quotes from various exchanges. To 5 show performance results over 30 days of test data, by indicator (2. Sharpe ratio; 3. Sortino ratio; 4. Max DD; 5. P&L-to-MAP), for the two baseline models , the Avellaneda-Stoikov model with genetically optimised parameters (AS-Gen) and the two Alpha-AS models.

I. How distant is the trader’s current inventory position is from the target position? (q)

First, the reward function can be tweaked to penalise drawdowns more directly. Other indicators, such as the Sortino ratio, can also be used in the reward function itself. Another approach is to explore risk management policies that include discretionary rules. Alternatively, experimenting with further layers to learn such policies autonomously may ultimately yield greater benefits, as indeed may simply altering the number of layers and neurons, or the loss functions, in the current architecture.

As we shall see in Section 4.2, the parameters for the direct Avellaneda-Stoikov model to which we compare the Alpha-AS model are fixed at a parameter tuning step once every 5 days of trading data. The two most important features for all three methods are the latest bid and ask quantities in the orderbook , followed closely by the bid and ask quantities immediately prior to the latest orderbook update and the latest best ask and bid prices . There is a general predominance of features corresponding to the latest orderbook movements (i.e., those denominated with low numerals, primarily 0 and 1). This may be a consequence of the markedly stochastic nature of market behaviour, which tends to limit the predictive power of any feature to proximate market movements. Hence the heightened importance of the latest market tick when determining the following action, even if the actor is beholden to take the same action repeatedly during the next 5 seconds, only re-evaluating the action-determining market features after said period has elapsed.

Background: The Avellaneda-Stoikov procedure

Therefore, the trader XRP will have the same risk as if he was using the symmetrical price strategy. These are additional parameters that you can reconfigure and use to customize the behavior of your strategy further. To change its settings, run the command config followed by the parameter name, e.g. config max_order_age. A closed-form solution for options with stochastic volatility with applications to bond and currency options. That is introduced with quadratic utility function and solved by providing a closed-form solution.

For instance, how are market prices (or actually differences to the mid-price) truncated to the interval [-1,1]? Are they scaled by some scaling parameter beforehand – and what data is this parameter estimated from ? If not, how much data is lost by only using the price differences with absolute values smaller than 1?

It can be also seen that the inventory of the trader reverts to zero more quickly than the symmetric strategy and the standard deviation of the inventory is produced less in the strategy. In order to see the time evolution of the process for larger inventory bounds. This part intends to show the numerical experiments and the behaviour of the market maker under the results given in Sect. Similar to the proof of Proposition2, the optimal spreads can be found by the first order optimality conditions. For the case of exponential utility function, now we explore the results of optimal controls obtained by solving the HJB Eq.

The combination of the choice of one from among four available for γ, with the choice of one among five values for the skew, consequently results in 20 possible actions for the agent to choose from, each being a distinct (γ, skew) pair. We chose a discrete action space for our experiment to apply RL to manipulate AS-related parameters, aiming keep the algorithm as simple and quickly trainable as possible. A continuous action space, as the one used to choose spread values in , may possibly perform better, but the algorithm would be more complex and the training time greater. With the risk aversion parameter, you tell the bot how much inventory risk you want to take.

The https://www.beaxy.com/ also places one bid and one ask order in response to every tick. Once every 5 seconds, the agent records the asymmetric dampened P&L it has obtained as its reward for placing these bid and ask orders during the latest 5-second time step. Based on the market state and the agent’s private indicators (i.e., its latest inventory levels and rewards), a prediction neural network outputs an action to take. As defined above, this action consists in setting the value of the risk aversion parameter, γ, in the Avellaneda-Stoikov formula to calculate the bid and ask prices, and the skew to be applied to these. The agent will place orders at the resulting skewed bid and ask prices, once every market tick during the next 5-second time step.

Two very important aspects in automated storage and retrieval systems (AS/RS) are productivity and maintenance costs. In the literature, as in industry, it is very difficult to find a solution that guarantees satisfactory results in terms of both. Moreover, all the solutions that the scientific community has proposed are static, i.e., the system’s behavior does not change as boundary conditions change. Many simulation runs’ results indicated that the DOF ensures a throughput time aligned to that of the benchmarks, but without needing to reorganize the stock during nonworking shifts.

Optimal dealer pricing under transactions and return uncertainty. Risk metrics and fine tuning of high frequency trading strategies. That is introduced by Avellaneda and Stoikov and handled by quadratic approximation approach.. And for the stock price dynamics which are provided in each model definition.

One of the most active areas of research in avellaneda & stoikovic trading is, broadly, the application of machine learning algorithms to derive trading decisions based on underlying trends in the volatile and hard to predict activity of securities markets. Machine learning approaches have been explored to obtain dynamic limit order placement strategies that attempt to adapt in real time to changing market conditions. As regards market making, the AS algorithm, or versions of it , have been used as benchmarks against which to measure the improved performance of the machine learning algorithms proposed, either working with simulated data or in backtests with real data.

We relied on random forests to filter state-defining features based on their importance according to three indicators. Various techniques are worth exploring in future work for this purpose, such as PCA, Autoencoders, Shapley values or Cluster Feature Importance . Other modifications to the neural network architectures presented here may prove advantageous. We mention neuroevolution to train the neural network using genetic algorithms and adversarial networks to improve the robustness of the market making algorithm. Reinforcement learning algorithms have been shown to be well-suited for use in high frequency trading contexts [16, 24–26, 37, 45, 46], which require low latency in placing orders together with a dynamic logic that is able to adapt to a rapidly changing environment. In the literature, reinforcement learning approaches to market making typically employ models that act directly on the agent’s order prices, without taking advantage of knowledge we may have of market behaviour or indeed findings in market-making theory.

Section 5 describes the experimental setup for backtests that were performed on our RL models, the Gen-AS model and two simple baselines. The results obtained from these tests are discussed in Section 6. The concluding Section 7 summarises the approach and findings, and outlines ideas for model improvement.