Guide

Robo advisors have three report types that are sent to each advisor designated channel.

  • Strategy Indicator reports

  • ŸRA Summary reports

  • Advisor state reports

All Ÿ Robo-Advisors generate the same type of reports except for Ÿ Macro.

Strategy Indicator

Each strategy indicator report has three sections:

  1. Graph: visualize the strategy and indicate whether it was it was Buy or Sell signal.

  2. Strategy: explains how to interpret the data presented in the report and provides “Buy” or “Sell” signal.

  3. 7 days Residuals data (undervalued/overvalued)

  1. Graph: The green and red dots represent the buy and sell signals, and the line connecting them shows the actual price of Bitcoin at that time.

The bottom of the graph has the residual values throughout the presented period. This helps paint a picture of current market trends and provides insight into how the residuals' deviated throughout the period.

  1. Strategy: This section explains how we calculate 'Residual' and how to understand if it's overvalued or undervalued. We also show you when it's considered a good time to buy or sell. (1.5x deviation positive or negative). If the deviation is greater than 1.5x deviation then it is a buy signal, if the deviation is less than -1.5x deviation, then it is a sell signal.

"Residual" is the difference between the price the model thinks BTC should have been & the actual price of BTC at that time.

  1. 7 days Residuals data: This section gives users the residual values for the last 7 days. The color-coded circles show if the values are under or overvalued. Red means it's overvalued (Sell), and Green means it's undervalued (Buy). The percentage tells you how much the residuals differ.

Example with Interpretation

Looking at the example above, we can see that

  • Within the past seven days there have been 7 "signals" i.e 1.5x deviations (positive or negative).

  • The price of BTC is shown to be undervalued with 3 days (red circles) where it was overvalued in 4 days (green circles).

  • The graph helps visualize BTC price and the buy/sell signals (the dots).

ŸRA Summary Reports

Each YRA Summary report contains 3 sections:

  1. Graph: shows the spread of predicated versus actual returns on different days.

  2. Strategy Stats: Provides various stats around the strategy deployed, such as the return, winrate & avg_win.

  3. Model Stats: Provides some details about the specific model being used.

  1. Graph: This shows the spread of predicted versus actual returns on different days. Red represents values from last week. Please note, that the colors used above don't indicate whether the price impact is positive or negative.

  2. Strat Stats: The strategy is to buy if the residual is above 1.5 standard deviations and sell if it's below 1.5 standard deviations. The stats are based on a dollar-cost averaging approach, where 5% of the total balance is bought or sold with each signal.

  • Return: The total return on investment during the test period

  • Winrate: The percentage of successful trades out of the total trades

  • AVG_win & AVG_loss: The average gain or loss for the strategy

  • Sharpe: Measures how much return you're getting for the risk you take. Read More on Sharpe Ratio

  • Sortino: Helps you see how much money you might make compared to the chance of losing it. It only considers the bad risks, like losing money. A higher Sortino Ratio suggests that the investment is safer. Read More on Sortino Ratio

  1. Model Stats: We’re using a model called "Ordinary Least Squares regression" or "ols_regression."

  • MSE (Mean Squared Error): This tells you how accurate the model is.

  • R2 (R Squared): This shows how well the model can predict the outcome based on the inputs.

  • Datapoints & Features: Features are the input variables used to train or test the model, and each datapoint includes several features.

Example with interpretation

The graph is a visualization of the strategy utilized by this model. Here we can note the deviations throughout & compare them to the previous week's stats.

Below that we get a look into the actual stats behind the strategy, using the example above, we can note that the model on average had a 2.46% return per trade. We also can see that using this strategy, the model has had a 100% winrate with its trades during this specific period. One can also note that on average throughout all trades, this strat has had a 12.03% win rate with not enough recorded data for an average loss rate to be calculated. Then we come across the Sharpe & Sortino values, our Sortino is rather high at 1.53 which could indicate the safety of said investment and our Sharpe is pointing to getting a positive return on the trade.

State Reports

Each Advisor State Report contains 3 sections:

  1. Basic Info: Details around the date & time of the last report/update

  2. Model Output: This section compares the model's expected price with the actual price, highlighting the residual value and indicating whether the current price is undervalued (green) or overvalued (red).

  3. Features: This section identifies market patterns using various data points to predict expected prices, showing feature values as positive (green), negative (red), or neutral (white).

  1. Basic Info: Contains data around the current date & the previous time of an update/report

  2. Model Output: This section provides a breakdown of what the model expected, the actual price and the different between the two, i.e the Residual value. We then present how overvalued or undervalued the current price is, this is given by taking the expected value - the actual value.

    1. Green: undervalued

    2. Red: Overvalued

  3. Features: This section provides you with some of the features used to predict the expected price on a given day. This section uses various data points to identify market patterns.

    1. Green: Positive Feature value

    2. Red: Negative Feature value

    3. White: Neutral feature value

Example with interpretation

In the example above, we are using ETF State Report. The model expected the price of BTC to be 1.5% overvalued, however the actual value of BTC was 2.15% overvalued. By minusing the Expected ($65104.33) by the Actual ($64674.01) we get an undervalued Residual of 0.65% ($431). Thus the model believes that the price of BTC at the moment is undervalued by exactly $431 (0.65%).

For the features section, we are presented with numerous leading BTC ETF's & a TOTAL value representing all of them combined. One can interpret that the model is indicating that the combined flow of money into the ETFs has been less than expected & we can note the flows in specific ETFs (GBTC & IBIT), etc. The TOTAL_PD provides insight into the total Premium/Discount that the specific or TOTAL ETF has.

The above applies to all our Ÿ Robo-advisors.

Ÿ Macro Reports

Ÿ Macro has three report types that are sent to Macro Advisor designated channel.

  • Upcoming Events Report

  • Events Summary

  • The Event Report

Upcoming Events report is updated in real-time & pinned at the top of the Ÿ Macro channel. It contains the following 3 sections:

  1. Event Name: Displays the name of the event that has occurred

  2. Time & Date: Provides for the Date & Time of the report

  3. Previous & Consensus Forecast values for economic indicator associated with the event

Event Summary is updated in real-time & pinned at the top of the Ÿ Macro channel. It contains the following 3 sections:

  1. Graph: Shows distribution of Target v Total Values

  2. Class Rules: Provides a breakdown of how the class rules work - Green is displayed when the value is higher than 0 - Red is displayed when the value is lower than 0 - White is displayed when the value is 0

  3. Events: Provides a list of the various economic events & their average impact on the price of BTC

The Event Report contains 5 sections:

  1. Graph: shows the distribution of Returns. neutral versus total returns on different days.

  2. Event: Details around the event, data & time.

  3. Actual Data: Provides various stats around the strategy deployed, such as the return, winrate & avg_win.

  4. Features: This section provides

  5. Model Stats: Provides some details about the specific model being used.

  1. Graph: Shows the distribution of Returns around this specific event (Core CPI MoM). Visualizing the data presented in the table below it

  2. Basic Info: This section provides basic information about the specific event specifically the name of the event, the data & time the event took place

  3. Actual Data: Provides stats around Previous, Forecasted & Actual value. Indicating a positive or negative change with Green or Red colors.

  4. Features: Delta calculation based on the input variables used to train or test the model. the actual and forecast delta value with error margin and performance classification (Positive, negative, neutral). (e.g., "negative" when the error is negative - Actual Value is less than the Forecast) (e.g., "Positive" when the error is positive - Actual value is higher than the Forecast) (e.g., "Neutral" when the error is 0.0- Actual value is the same as the Forecast)

  5. Stats: Statistical predictions (Mean, Median, Minimum and Max) for a 7 day time horizon to reveal market behavior in response to the class result for each of the Y Macro events reported (Positive, negative, neutral).

Example with Interpretation

One can interpret from the data above that the event's outcome was expected thus we are seeing a neutral response by the market to this news. We expected a 0.2 change and were met with the same. If we look at the Features section, we are told that the event is seen as Neutral by the model.

Since the event outcome was as expected, the market's reaction has been neutral with barely any volatile movements happening. We can also note that the average return for BTC for the 7 days is only +0.56% which suggests a slightly positive impact. Paired with the Median stat, we note that the impact will indeed be neutral & slight at best. We had some volatile swings from +12.66% to -12.75% but can note that we settled on a neutral +0.56% impact on BTC.

V2 Robo-Advisor Tutorial

PreviousŸ Robo-AdvisorsNextTimeline

Last updated