Testing Bitcoin Market Efficiency with Neural Networks and Cross-Asset Inputs
Summary
The study tests the weak form of the efficient market hypothesis for Bitcoin with a feedforward neural network. It asks whether adding returns or related features from other assets improves prediction, alongside Bitcoin returns. The additional inputs include equity indices, the EUR/USD exchange rate, a US Treasury yield, and an index of gold and silver producers. The study compares training performance with test-set prediction and considers performance relative to buy and hold.
Training accuracy rises from 54.6% with one feature to 61% with six, but the added features do not improve test-set accuracy. One feature set partly outperforms buy and hold, while adding another feature reduces that performance. The authors therefore find that their model and selected inputs do not reveal exploitable weak-form inefficiency during the sample period, while qualifying this as a partial conclusion. The result depends on the chosen period, features, architecture, and hyperparameters; it does not establish that Bitcoin is always efficient or rule out other methods.
Key ideas
- The study tests Bitcoin's weak-form efficiency using a feedforward neural network.
- It compares Bitcoin returns alone with inputs from equities, currencies, Treasury yields, and a metals-related index.
- Training accuracy improves as features are added, but test accuracy does not.
- One feature set partly outperforms buy and hold, with weaker performance after another feature is added.
- The efficiency conclusion is limited to the studied sample, inputs, and model configuration.
Tags
Full text
# The Efficient Market Hypothesis for Bitcoin in the context of neural networks # The Efficient Market Hypothesis for Bitcoin in the context of neural networks This study examines the weak form of the efficient market hypothesis for Bitcoin using a feedforward neural network. Due to the increasing popularity of cryptocurrencies in recent years, the question has arisen, as to whether market inefficiencies could be exploited in Bitcoin. Several studies we refer to here discuss this topic in the context of Bitcoin using either statistical tests or machine learning methods, mostly relying exclusively on data from Bitcoin itself. Results regarding market efficiency vary from study to study. In this study, however, the focus is on applying various asset-related input features in a neural network. The aim is to investigate whether the prediction accuracy improves when adding equity stock indices (S&P 500, Russell 2000), currencies (EURUSD), 10 Year US Treasury Note Yield as well as Gold&Silver producers index (XAU), in addition to using Bitcoin returns as input feature. As expected, the results show that more features lead to higher training performance from 54.6% prediction accuracy with one feature to 61% with six features. On the test set, we observe that with our neural network methodology, adding additional asset classes, no increase in prediction accuracy is achieved. One feature set is able to partially outperform a buy-and-hold strategy, but the performance drops again as soon as another feature is added. This leads us to the partial conclusion that weak market inefficiencies for Bitcoin cannot be detected using neural networks and the given asset classes as input. Therefore, based on this study, we find evidence that the Bitcoin market is efficient in the sense of the efficient market hypothesis during the sample period. We encourage further research in this area, as much depends on the sample period chosen, the input features, the model architecture, and the hyperparameters.
Shown in full with attribution under the source's licence. Licence: abstract CC0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.