The problem with DCA in a bear market
Dollar Cost Averaging is one of the most popular trading strategies in crypto. The idea is simple: buy the dip. Place a base order, and if the price drops, buy more at lower prices with safety orders. Each purchase lowers your average entry. When the price bounces, even a small rebound triggers take profit.
In a bull market, this is essentially free money. Every dip is a buying opportunity, and every bounce is a payday. DCA bots print deals.
But in a bear market, DCA is a slow bleed. The price drops, you buy the dip. It drops more, you buy more. It keeps dropping. Your safety orders fill, your average entry is still above the market, and the bounce never comes. Stop-loss fires. You eat the loss. Repeat.
We backtested standard DCA across a full year of data covering both bull and bear regimes. The results were illuminating:
ETH barely broke even. $21 profit on a year of trading. And that 16.9% drawdown? That's real money at risk, eating into your capital while the bot grinds through losing deals in the bear phase.
What if the bot knew which way the market was going?
This is the question that led to Smart DCA.
Standard DCA only goes long. It buys dips and sells bounces. That's the only move it has. When the market is falling, it's playing offense in the wrong direction.
Smart DCA adds a trend filter — a proprietary signal that reads the macro direction of the market. When the trend is up, Smart DCA goes long, just like regular DCA. But when the trend flips bearish, Smart DCA reverses: it goes short.
Short DCA is the mirror image:
- Base order shorts at the current price
- Safety orders trigger when the price rises (averaging up your short entry)
- Take profit fires when the price drops below your average entry
In a bear market, this is exactly what you want. Every rally is a shorting opportunity. Every dump is a payday. The bot profits from falling prices instead of bleeding against them.
The numbers
We backtested Smart DCA against regular DCA across a full year of BTC and ETH data — a period that included a massive bull run followed by a sustained bear market. Same parameters, same risk. The only difference: the trend filter.
| Asset | Regular DCA | Smart DCA | Improvement |
|---|---|---|---|
| BTC/USDT | +$86.54 | +$149.83 | +73% |
| ETH/USDT | +$20.73 | +$462.85 | +2,133% |
On BTC, Smart DCA earned 73% more than regular DCA. On ETH, the improvement was over twenty times. ETH went from a barely-profitable grind to a seriously impressive return — because the bot was shorting through the bear phase instead of bleeding through it.
Win rate and drawdown
ETH's win rate jumped from 87% to 93%. More deals closed in profit. The drawdown stayed the same while returns went up 22x.
How it works (without the secret sauce)
We won't reveal the exact indicator or parameters — that's our edge. But here's the principle:
- Trend detection: A moving average crossover signal classifies the market as bullish or bearish. This is a well-known concept in technical analysis — the specific implementation is what matters.
- Direction locking: When the bot starts a new deal, it checks the current trend direction and locks it for that deal. No mid-deal direction changes. No whipsawing.
- Automatic flipping: Between deals, the direction is recalculated. If the market has shifted from bull to bear (or vice versa), the next deal opens in the new direction.
- Full DCA mechanics: Everything else is the same as regular DCA. Same base order, same safety orders, same take profit %, same stop-loss protection. The only difference is which way the deals point.
The best part: you don't need to decide the direction yourself. The trend filter handles it. You just set your DCA parameters and let the bot adapt to whatever the market does.
When Smart DCA isn't the right choice
We're not going to pretend Smart DCA is magic. It performs best on large-cap assets with clear macro trends — BTC and ETH being the prime examples. These assets have long sustained trends that the filter can lock onto.
On highly volatile, choppy altcoins (like SOL or meme coins), the trend signal can whipsaw. In those cases, regular DCA actually outperforms because it doesn't try to predict direction — it just buys every dip regardless.
That's why we built the backtester. Before you deploy Smart DCA on any pair, you can test it against historical data and see exactly how it would have performed. No guessing, no faith-based trading. Just numbers.
Backtest it yourself
Smart DCA is available right now on all myRaijin accounts, including free tier. Run a backtest on BTC or ETH over the past 6 months. Compare it to regular DCA. See the deal log. See the equity curve. See the difference.
The backtest doesn't lie. If Smart DCA works for your pair and your timeframe, the numbers will tell you. And if it doesn't, regular DCA is right there as a fallback.
We built this because we trade our own money with these bots. We got tired of watching DCA bleed through bear markets. Smart DCA is our answer. Now it's yours too.
← Back to blog