Recurring orders, often referred to as dollar-cost averaging, are a way to spread out an investment amount over time. This decreases exposure to price volatility and reduces time spent trying to time the market. It’s a way to passively invest in bitcoin in the long term.

River recently started offering zero-fee hourly recurring orders, in addition to the zero-fee daily, weekly, and monthly options we have had for a while.

This feature has led to two categories of questions:

1. How are we able to offer a zero-fee product? Is there a hidden cost in the price spread? You can read the explanation on our blog.
2. Is there a best time of the day or best day of the week to dollar-cost average bitcoin? And is there a best frequency for DCA?

Based on our research, we found that:

1. Historically, the daily high price happened within a 4-hour window 38.5% of the time, and the daily low price happened within that same 4-hour window 39.4% of the time. This pattern has shifted over time and the window has shrunk.
2. From 12-1 PM Eastern time, there is a one-hour window with over 3 times as many price bottoms than peaks in the past twelve months. This hour has a 4.37% theoretical advantage for daily recurring orders relative to any day’s average.
3. Mondays have historically had the highest odds of having the weekly low price relative to the weekly high price falling on this day. This day has a 14.36% theoretical advantage for weekly recurring orders relative to any week’s average.
4. The first and second days of each month have historically had the highest odds of having the lowest price relative to the monthly high price happening on those days. A monthly recurring order on these days gives a 6.83% and 3.73% theoretical advantage relative to any month’s average, while on the last three days of the month, there is a 3.11%, 6.83%, and 6.21% higher chance to buy a monthly high than a monthly low.
5. Due to the variance of the exact moments when price peaks and bottoms happen, the achievable practical advantage on timing a recurring order is around 1.2%.

You would think some of these findings are common knowledge among Bitcoin traders and investors, right? 

We asked around, and apparently, many of them don’t know. 

We will dive into each of these findings below.

1. Daily Highs and Lows in Bitcoin Price

We looked at bitcoin’s price since it first got one back in 2010. Each day, it has a highest and lowest point, the values and timing of which were recorded by CoinMarketCap and publicly made available in their dataset. The timing is only visible upon downloading.

We mapped out daily highs and lows by hour for 4860 days, and came across a clear pattern: the daily high price happened within a 4-hour window 38.5% of the time, and the daily low price happened within that same 4-hour window 39.4% of the time.

The first question that comes to mind is: why these specific hours?

Part of the answer is revealed when converting the timezone on the graph above to European and Asian timezones: This is when the highest number of traders can be awake.

For the US, the highest odds of the daily high and low price are in the early evening. For the EU around midnight, and for a large part of Asia, in the early morning. Besides this 4-hour window, the distribution of daily highs and lows is remarkably flat.

2. The Best Time to Daily DCA Bitcoin

The analysis above was done over the past 13 years, but do these findings still hold up in recent history? We had a look at the past 12 months, from November 2022 to October 2023, and found three changes in the pattern.

1. There is now an 8-hour window with significantly fewer high- and low price points.
2. Instead of a 4-hour window of clear outliers, there is now only one hour that stands out, which accounts for 14.75% of high price points and 15.03% of low price points. This is still statistically significant for traders and investors.
3. 12-1 PM Eastern time is a clear outlier with significantly more low price points than high price points. This translates to 5 PM London time, and 1 AM Hong Kong time.
   
   This makes it an appealing time for daily recurring orders to be executed, as there is a 6.28% chance of buying the daily low price, versus a 1.91% chance of buying the daily high price, the biggest gap on average across the year. This pattern may not hold up in the long term and should be tested from time to time, but it could help to stack a few more sats over time. Note that on River, daily recurring orders execute at the time they were first set up.

The question remains: why is the 4-hour window pattern changing? We found some context when looking at price volatility by hour.

Bitcoin Volume by Hour of the Day

In 2022, Coin Metrics published data on the share of volume traded per hour throughout the day. This provides an overview of when most money is being traded, which is not always the same as when the most price fluctuation is happening or the most profit can be made.

There is a clear trend of US mornings having the most volume on average. The dataset Coin Metrics used is international, so it was not skewed based on data from specific exchanges.

Mornings did not only have the most volume, but they did so at an increasing rate every quarter, while the volume in the evening dropped rather significantly. 

The evening hours historically represented the 4-hour window we raised earlier. It could be that volumes were reduced because the peaks were occurring less often, or that volumes shifting to US daytime lead to fewer peaks in the evening.

3. The Best Day to Weekly DCA Bitcoin

Similar to the best time of the day to DCA, we also found a weekly pattern.

Since 2010, Mondays have had the highest odds of having the weekly low price relative to the weekly high price falling on this day. This pattern holds up over the last 12 months.

Monday had a 14.36% theoretical advantage over the past year for weekly recurring orders relative to the average per week. Friday was also an outlier over the past year with a 10.94% theoretical advantage, but historically this advantage was only 2.15%.

The reason why those specific days stand out is likely related to the macro economy, which still operates on business days. As a result, both Monday and Friday are more likely to have volatility, as market actors react to news during the weekend. 

It is important to clarify that the statistical advantage to buy at lower price points does not equate an equal increase in profits, hence why we are referring to it as theoretical. 

A statistically higher chance of buying at a below-average price does not mean it always happens often enough. The timing during the day of when a recurring order executes has an impact on whether you get close to the daily low price or not.

To visualize this with an example:

If you had bought $10 worth of Bitcoin each week at 12PM Eastern time over the past five years, you would have invested $2,610, and made an estimated profit of:

That is only a $55.65 advantage, which is 1.2%, and thus drastically lower than the 14.36% statistical advantage. The reason is that while Monday may have a higher ratio of low price points to high price points, the low price points only affect recurring orders when they happen close to execution time, and in practice they will happen throughout various points of the day. Over a long enough period, there should be a more noticeable difference if the pattern holds up, but for now, there is significant variance still.

An example 1.2% more profit over five years will not be meaningful enough for many investors to act on, especially when it is not a given. At the same time, it is also reassuring for investors to know that the exact time at which they set up their recurring order won’t massively affect how much Bitcoin they accumulate.

4. The Best Day to Monthly DCA Bitcoin

For a monthly recurring order, there is a pattern as well when we look at the data since 2010, but it is not statistically significant enough to result in a meaningful impact.

An investor would have to DCA for many years to benefit from the 6.83% and 3.73% theoretical advantages on the first two days of the month, and it would take many years to validate whether the pattern still holds up. Out of the daily, weekly, and monthly patterns, this one seems the least impactful to act on.

On the other hand, the last three days of the month certainly seem worth avoiding. There is little drawback to setting up a recurring order on some of the earlier days in the month, to at least not have potentially higher odds of buying a monthly price peak.

We did not include a graph for the past twelve months here, as the sample size only consists of 12 high prices and 12 low prices, compared to 161 of each in the graph above.

5. Which DCA Frequency Is the Best?

There are many ways to analyze this question, and many factors that influence the decision.

1. Market conditions

Dollar-cost averaging performs differently in a bull market (continuously rising price over time), a bear market (continuously dropping price over time), and a relatively stable market.

2. Time period taken

With these analyses, it can be tempting to cherrypick a timeframe in which the results match your personal preference. However, past performance is not indicative of future results. Depending on during which period you start and how volatile markets are, outcomes may differ.

3. User goal

If it is your goal to accumulate as much bitcoin as possible, then a lower frequency for a recurring order has historically been a bit better.

If it is your goal to accumulate bitcoin, but always be able to convert it back to as close to the number of dollars you invested in the short term, then a higher frequency for a recurring order has historically been a bit better.

Conclusion

There are a few interesting patterns in bitcoin’s price that could lead to small advantages for recurring orders if a user is attempting to accumulate as much bitcoin as possible.

However, it's important to remember that past performance is not indicative of future results. The effectiveness of these strategies varies depending on market conditions and the time frame that is being considered. In general, the exact timing of a recurring order will not have a material impact on many investors.

The recurring order frequency can have an impact: A higher frequency allows an investor to remain closer to the dollar amount they invested in case they need to exchange their investment back at any time. A lower frequency can accumulate a bit more bitcoin in the long run.

The principles of recurring orders remain regardless of which frequency or timing is chosen:

• Lower your exposure to price volatility.
• Reduce your time invested to try to time the market.
• Turn investing in bitcoin into a long-term, passive activity.

Disclaimer

This report was prepared for informational purposes only and does not represent investment advice of any kind.

River Financial, Inc. does not provide tax, legal, investment, or accounting advice, and this report should not be relied on or construed as such. We recommend performing your own analysis and seeking professional advice before making any financial decisions.

Information contained in the report is based on either our own data or external sources we consider to be reliable. We cannot guarantee the accuracy or completeness of all data.

River Financial, Inc. shall have no liability whatsoever for any actions taken or decisions made as a consequence of the information in this report, including any expenses, losses, or damages, whether direct or indirect.

The contents of this report are the property of River Financial, Inc. and may not be duplicated or distributed without the prior written consent of River Financial, Inc.