AQTIS - Trend Following (Long)

We are excited to announce the release of our first performance report for our trend-following trading strategy.

Within this report, we illuminate the implementation of trend-following strategies across diverse assets and time frames, providing a comprehensive showcase. Delving into the potential returns, we demonstrate the impressive capability of AQTIS to generate substantial gains.

Introduction

AQTIS mission

AQTIS is on a mission to create a diverse system made up of different trading strategies that allow AQTIS to deliver a yield, whatever the weather. To achieve this, over the last year, we have developed three main strategies to create a diverse quant technology stack.

Those strategies have been paired with our own machine learning and AI algorithms to select the best strategies for the appropriate market conditions. Not every strategy works well in every market, requiring the ability to adjust strategies according to market conditions.

The results of some of our other strategies have been published, and can all be found here.

In this report, we publish the result of our Trend Following Strategy.

Trend Following

What exactly is a trend?

The trend phase of the market is the opposite of a sideways-moving market.
A trending market creates a series of new highs and lows and moves in a certain direction, either up or down.
According to most sources, the market is in a trending phase only 20%-30% of the time. Yet, it is precisely during these periods of volatility that the greatest profits can be generated.

A classic method to identify a new trend is by using moving averages.
This can be done using a single SMA or EMA, where each time the price crosses the SMA or EMA, it indicates a change in market direction.
Alternatively, two moving averages with different periods – one shorter than the other – can be used. When these SMAs or EMAs intersect, it confirms a shift in market direction.

How do we identify a trend?

At AQTIS, we adopt a different approach. We focus on assessing the strength or weakness of the momentum.

Unlike in classical Technical Analysis where traders use lagging indicators like RSI and Moving Averages, our Quant Tech looks at orderbook ratios, significant changes in Open Interest, volume and other real-time indicators to identify changes in momentum and assesses the probability of a directional move.

Also, it allows us to rank coins and assign the optimal size to every asset.
When we detect extreme momentum, we “bet” that this momentum will lead to the formation of a new trend, and we capitalize on this movement

The Model

Once our system detects extreme momentum in one direction, knowing the optimal Take Profit area and the invalidation conditions for each pair (Stop Loss), it creates an order with the optimal position size for our portfolio.

Two key components are implemented in the trend-following system:

BTC - Multi Timeframe Analysis

The robustness of our trend-following strategy renders it consistently profitable across various assets and timeframes.

To illustrate, we will showcase the strategy's performance and highlight the distinctions in entry points for BTC/USDT across 1D, 4H, and 1H timeframes.

BTC/USDT - Trend Following - Long Only - 1D

BTC/USDT - Trend Following - Long Only - 4H

BTC/USDT - Trend Following - Long Only - 1H

As evident, the strategy proves highly profitable over the long term. Thanks to the regime filter and exit conditions, once favorable returns are achieved, they are not given back to the market. Of course, an asset with clear trends is necessary to profit from this system, and cryptocurrencies are well-suited due to their directional nature.

However, as shown in the upper charts, reducing the timeframe and increasing the number of trades does not necessarily lead to higher returns. Due to the costs of transactions (fees, slippage), increasing the number of trades can be counterproductive. Additionally, we may exit a trade too early, cutting off potential future profits due to movements in shorter timeframes.

Holding performance

The year 2023 was a good one for cryptocurrencies, but significant price increases occurred mainly in two key periods: January and February, and from October to December.
The rest of the year, the movement was not as clear.

This can be observed in our 'spaghetti' chart plotting the 20 assets we selected in summer.
The chart shows significant fluctuations, with both steep rises and sharp falls, indicating that by September, our position would be either void or below our initial 2023 investment.

You can view the average performance and the median return in the lower subplot.

Trend Following Performance - Long 1H

Our goal with the trend-following system is straightforward: when we detect a bullish movement, we enter the trade and exit when the momentum fades. In the meantime, the aim is to avoid losing money when market conditions are not conducive to an upward trend.

This approach allows us to capitalize on the bullish trends that so distinctly define the cryptocurrency market, while achieving lower volatility in our portfolio.
The trade-off for reducing this volatility is that we won't ride every trend to its end. However, this strategy ensures we secure profits before the market reverses.

Key metrics and numbers

Here you have access to the backtesting data for each asset. You can view all the parameters of the backtest.

Conclusions

In conclusion, we can highlight the following:

What's next?