Setting the scene

Imagine you had a wise friend who watched the markets every day. They made links between different streams of information and they compiled this information into a ‘gut feel’.

They didn’t fall for overconfident predictions… they knew the markets are unpredictable

But through the noise they tracked trends

When shit coins were flying, they grew wary and when no one was talking about cryptocurrency, they became greedy.

They were always bullish on crypto, so they were always buying…but they moderated their buy amount according to their gut feel.

They bought less when things felt risky. They bought more when things felt safe.

‍

Building that friend

At CALC, we thought that everyone could do with a friend like this.   A friend that could take a risk-averse approach to invest, while still extracting maximum value given the bountiful information available to them.

We wondered if the powerful pattern recognition skills of machine learning algorithms could match those of your friend.

And what we came up with is DCA+, your set-and-forget robo friend that watches the market for you, 24/7, and guides your decision-making. So you can spend more time doing the things you love, while still generating maximum gains.

‍

How does DCA+ work?

It minimizes risk compared to lump-sum investing by spreading out buys through time

It maximizes gains compared to traditional DCA by moderating the buy amount according to market conditions.

How does it moderate the buy amount? It moderates buy amount according to a machine-learning powered risk analysis framework, such that the weighting of risk associated with each buy-interval is equal across all buys – we call this ‘risk averaging’

Risk = high, buy less

Risk = low, buy more

Unlike other alternatives to traditional DCA like ‘Value Averaging’, where the algorithm sometimes instructs enormous buys, DCA+ buy amounts are bound by confidence in the risk assessment.

You’ll never buy more than ~2.5x what you otherwise would using a traditional DCA approach. And inversely, you’ll never buy less than ~0.4x. This way, errors in the machine learning risk assessment are minimized, to keep you safe in the market

…And how does this strategy perform?

A case study:

If you were to have begun to DCA+ into $ATOM on the XX:XX:XXXX aimed at accumulating over the next 180 days, you would’ve accumulated XX more ATOM, equating to a return of XX% more in USD.

You can see that the buying strategy was more aggressive early in the DCA window, indicating that risk was (correctly) perceived as relatively low by the algorithm.

It is clear that DCA+ strategies really shine in the period before the bull market kicks off. More capital allocated before the bull = greater returns. However, when the bull market was in full swing and risk was perceived as higher, less capital was allocated, resulting in less bought at the top, and still improved returns compared to traditional DCA.

DCA+ is designed to work in all stages of the market cycle, it is truly set and forget.

‍

Does this only work for $ATOM?

In keeping with the goals of DCA, we wanted to keep a conservative approach to risk and thus tailored our algorithm to that of the price movement of large-cap currencies. You’ll be able to DCA+ into $ETH, $ATOM, $BNB, and $DOT at launch and eventually, nBTC when it arrives in the Cosmos.

‍

And what is the risk?

The traditional DCA strategy is great because it reduces your chances of getting rekt in extremely volatile markets. DCA+ manages to preserve the bones of this battle-tested risk mitigation strategy by keeping buy amount within bounds set by confidence in the machine learning assessments. How does it compare?

The lowest returns generated by both traditional DCA and DCA+ are essentially equivalent. Clearly, any increased risk associated with buying more is offset by buying less in high-risk times.

Our conclusion: more assets, same risk.

But you can make your own conclusions.