Current Adjusted Market Inefficiency Metric for the top 5 Cryptos by Market Cap

• Bitcoin (BTC): -1.240
• Ethereum (ETH): -0.465
• Binance Coin (BNB): -0.324
• Cardano (ADA): -0.294
• Solana (SOL): -0.377

AMIM is a value in the range of (-∞,1] with a negative value implying an efficient market, or in other words there is little autocorrelation within the price history time series.

These results agree with existing research which found BTC to currently be an efficient market.

What having the ability to quickly calculate AMIM means for placing automated trades.

We can now evaluate the market efficiency for hundreds of cryptocurrencies at once, and sort for cryptos that could be easier to forecast with existing predictive techniques.

Forecasting Accuracy Metric

To get preliminary data on the forecasting accuracy of the predictive model we are using, we split up the data set (price history time series) used to calculate market efficiency into a training and test set. We split up the data 80% training and 20% testing.

Below are the 18-day price forecasts made for the 5 cryptos AMIM was calculated for.

The below predictions and respective mean absolute percent errors are examples of how we would quantify the accuracy of our model.

Sprint 5 Goals

1. Calculate and sort the AMIM of as many cryptos as possible in the search of an inefficient market.
2. Forecast the price for inefficient markets and compare with efficient market forecasting.
3. Begin coming up with a standardized procedure so that we can utilize price forecast information to automatically place trades.