Skip to main content
Account

Table 2 Out-of-sample backtesting averaged returns (average value of PnL returns of multiple RL agent’s testing episodes) of the best PMM strategies and the benchmarks

From: Predictive Market Making via Machine Learning

Equity

\(PMM_{(MLP, w=0.80)}\) ($) strategy 1

\(PMM_{(LSTM, w=0.90)}\) ($) strategy 2

\(PMM_{(CNN, w=0.95)}\) ($) strategy 3

\({ PMM_{(perfect, w=0.50)}}\) ($)

RMM-Spooner ($)

AS model ($)

VOD

6.6356

4.3670

10.3290

14.4481

9.7375

-4.3579

AAL

-2.0023

-1.9823

-2.4106

-0.3446

-1.9563

-4.3562

GSK

-6.4512

-10.7045

-5.9494

-8.8139

-14.2216

-4.3849

MO

4.2870

-3.8337

8.5449

6.7650

7.3401

-4.4449

AMZN

3992.554

3885.496

3951.784

4176.199

3828.689

0.0132

WMT

-7.8498

-0.9850

-7.8599

0.5000

-4.5429

-4.6585

NVDA

-4.3581

-1.9429

1.2900

1.9002

-1.6871

0.2147

CVX

-18.7016

-13.5129

-20.179

-2.1177

-14.2416

-4.3447

UPS

-2.6849

-3.6351

-1.1051

-0.955

-1.775

-4.4126

TXN

-16.4483

-12.575

-14.934

-5.5586

-15.7651

-3.1253