Cash forecasting is important because it provides visibility into all the variables that could affect future liquidity. The most important aspects of cash forecasting are the assumptions used to create the forecast. We designed forecasting in Trovata to give our customers an Excel alternative to store and manage those assumptions, as well as provide automated reporting of actual cash flows for cleaner variance reporting.

After creating the forecast and monitoring for variances, you will learn new insights about your cash flows. Also, by centralizing your Forecasts inside Trovata, you can give your team access to own and control their data, all while avoiding Excel errors, file corruption, and manually reporting cash flow actuals.

To start forecasting in Trovata, you first need to define the purpose of the forecast and your intended audience.

Some examples may include: general liquidity planning, anticipating liquidity choke points due to growth, investments, black swan events, or simply to communicate the health of the company to key stakeholders.

Here is a summary of how to create Forecasts inside Trovata:

Based on your goals and the intended audience, you will first create Tags inside Trovata. These tags need to match the categories of cash flows that are important for your audience to see when reviewing your forecast (i.e. AR, AP, Payroll, Taxes, etc.). These tags will also be used to track actuals for variance reporting.

Once Tags are created, you'll navigate to the Forecasts module inside Trovata to begin building Forecast data streams. These data streams represent the cash flow activity within your forecast. A stream can represent a cash inflow or outflow.

Within Trovata, Data Streams are the main building blocks of each forecast. There are various methods for creating data streams. These methods can be used interchangeably within a forecast, meaning you can build a forecast using multiple assumptions methods.

Data streams are the building blocks of each forecast in Trovata.

If you would like to learn how to create a new data stream, keep reading.

Combinations of the methods described below can be used within a forecast, meaning you can build Forecasts using multiple stream creation methods.

There are two stream types:

Data Streams utilize a direct entry method to managed cash forecast values. Data streams can be populated using machine learning, repeating historical data, or manual entry.

Invoice Streams use open receivable and open payable invoices from an ERP system to predict future cash forecast values.

The following sections will explain how to create each stream type and explain the unique features available for each stream creation method.

To create a stream, access the Data Stream library by clicking Data Stream in the Trovata menu.

To create a new stream, click New Stream in the top-right corner.

Machine Learning (ML) based streams are produced by providing historical transaction data to multiple forecasting models.

Use Cases:

To create a Machine Learning (ML) based streams click New Stream → Data Stream.

Input the Stream Name and choose a Currency for this Data Stream.

Click Next.

Choose the Analysis option and click Next.

Using the Search bar or Filtering menu, Data Sources can be found by searching the Trovata database of bank transactions.

After choosing a Tag (or group of Tags) as the Forecasted Data Source, it is visible for reference under the Search bar.

By default, the previous 13 weeks of transaction data will be used for the Machine Learning analysis.

Click the Date Range picker to choose a specific period of time to use for the Machine Learning analysis methods.

The transactions result from the search and date range option appear below the search bar.

Click Next: Forecast Method → Machine Learning (ML).

As a result, Trovata’s Machine Learning models will analyze the Forecasted Data Source and provide four (4) scenario models.

Minimum and maximum values for the Machine Learning results can be applied to restrict the result values.

Click Done.

Review the data configuration and click Next.

For variance reporting, bank actuals need to be reported against the forecasted stream data.

Review the selection for how to manage actuals data.

If the Actuals data configuration needs adjustment, click the three dots to the right, otherwise click Next.

Once a review is complete, click Confirm to finalized the creation of this stream.

Upon stream creation, a graph and data table of the forecast results are shown.

Repeat History data stream uses a selection of historical bank data (i.e. 13 weeks, 12 months, etc.) and copies it forward.

Use Cases:

To create a Repeat History based streams click New Stream → Data Stream.

Input the Stream Name and choose the Currency.

Click Next.

Choose the Analysis option and click Next.

Using the Search bar or Filtering menu, Data Sources can be found by searching the database of bank transactions in Trovata.

After choosing a Tag (or group of Tags) as the Forecasted Data Source, it is visible for reference under the Search bar.

By default, the previous 13 weeks of transaction data will be used for Repeat History.

Click the Date Range picker to choose a specific period of time to use for the Repeat History method.

The transactions result from the search and date range option appear in the graph below, followed by the data table.

Click Next: Forecast Method → Repeat History.

The date in which the Repeat History value begins in the forecasted data is determined by the Import Start Date.

Click Import Start Date and choose from the date picker to adjust.

The historical values can be copied exactly, or a change to be applied to adjust the forecasted values by a percentage or dollar amount.

Click Amount ($) to define a change to the historical data by a dollar amount or percentage (i.e. Amount ($) or Percentage (%) from the dropdown list).

Click Change field to input the value of the dollar or percentage change.

Click # 1 field here to determine the number of times that the historical data should be repeated in the forecasted data stream.

Click Done.

Review the data configuration and click Next.

For variance reporting, bank actuals need to be reported against the forecasted stream data.

Review the selection for how to manage actuals data.

If the Actuals data configuration needs adjustment, click the three dots to the right, otherwise click Next.

Once a review is complete, click Confirm to finalized the creation of this stream.

Upon stream creation, a graph and data table of the forecast results are shown.

Manual Entry data stream constructed entirely by one-off user inputs and/or a CSV file upload.

Manual data can be entered into the forecasted data stream in the following ways:

Use Cases:

To create a Manual Entry based streams click New Stream → Data Stream.

A New Data Stream creation wizard will open. Begin with Basic Info by entering the Stream Name and choosing the Currency for this Data Stream.

Click Next.

Choose the Manual option and click Next.

Much like a spreadsheet, the Forecast Value column will accept forecast values directly.

Click into the Forecasted Value cell of the row associated with the forecast Date and Type the value as a positive (receipt of cash) or negative number (disbursement of cash).

Again, like a spreadsheet, the Forecast Value cells can be populated using the Copy and Paste keyboard shortcuts

Using the Analysis method is the exact same as using the Repeat History method described earlier in this guide.

Click Input → Analysis to access the configuration wizard described previously within the Repeat History section of this guide.

Auto-Entry provides tools create forecast assumptions By Cadence, using Repeat Data to carrying forward existing data within the stream, or to adjust Weekend Data.

Click Input → Auto-Entry to access the Auto-Entry tools.

From a particular date onwards, forecast values can be replaced or added to an existing amount and repeated at specific cadence and for a number of periods.

Choose a Start Date.

Select Entry Type, Replace existing or Add to existing.

Input Amount.

Choose a Cadence from the list: Daily, Weekly, Bi-Weekly, or Monthly.

Input the number of Periods that amount should be repeated at the selected Cadence.

Click Submit to save.

The Repeat Data tool will copy a section of forecast data that has already been entered in a stream and copy it forward.

Choose a Date Range to select a range of existing forecast values to repeat.

Similar to the Repeat History method describe in the section above, a Change by Amount ($) or Percentage (%) can be applied to the daily forecast values.

Pick the Start Date that the Repeat Data selection should begin.

Input a value for the Times Repeated field.

Many companies do not have cash activity over the weekend. In these cases, weekend values that are populated in the forecast stream using any of the methods above can be adjusted.

Options include:

Choose a Start Date, which will limited to every Saturday in the date picker.

Input the number of Periods that weekend data should be adjusted.

Pick the Weekend Data option from the dropdown.

In addition to copying and pasting data from spreadsheet, .csv files can be uploaded.

To upload a .csv file, click CSV Upload and choose the appropriate file from your device.

Upload .csv files must contain columns for Date and Value.
​

Click Done.

Review the data configuration and click Next.

For variance reporting, bank actuals need to be reported against the forecasted stream data.

Review the selection for how to manage actuals data.

To choose an existing Tag to managed Actuals data, click Analysis.

Using the Search bar or Filtering menu, Data Sources can be found by searching the database of bank transactions in Trovata.

After choosing a Tag (or group of Tags) as the Forecasted Data Source, it is visible for reference under the Search bar.

Click Done.

Once actuals data configuration is complete, review and click Next.

Once a review is complete, click Confirm to finalized the creation of this stream.

Upon stream creation, a graph and data table of the forecast results are shown.

Invoice Streams use open receivable and open payable invoices from your ERP system to predict future cash flows.

All invoices data available to use in Trovata can be found within the Open Invoices submenu.

To create an Invoice Stream, click the Data Streams menu item. Then click New Stream → Invoice Stream.

Input the Stream Name, Currency, and click Next.

Click Configure data source.

From this page, the Search bar can be used to find specific invoice data, like Customer/Vendor, ERP, or Invoice Type.

Alternatively, the Filter to the right can be used.

Once the invoice data is defined, a Max Age can be set.

Click Done.

Click Next.

Review and click Confirm.

The main ingredients we need to assemble a forecast are:

To create the Forecast, choose New Forecast in the top right hand corner of the screen.

Give your Forecast a name, choose the base currency for the Forecast, as well as the default Cadence in which you wish to view your forecast output in - e.g. daily, weekly, monthly, or quarterly. You also have the ability here to select your rounding treatment. (No Rounding is the default method and most commonly used).

All of these options can be changed later on.

Next, choose the bank accounts that correspond with the cash flows you are forecasting. This may be all accounts if you are forecasting for the company as a whole, or just a select few if you are only forecasting for a specific operating account or an entity.

Choose the data sources for the forecast. These are most likely going to be the data streams you created earlier. You may notice that you can also choose other forecasts as data sources - giving you the ability to create a Forecast of Forecasts, or also know as “Forecast Roll-Ups” (see below).

As an option, you can apply Global Factors to your forecast. This feature is popular when creating multiple forecast scenarios or for more advanced forecasting.

Finally, you will be able to review all the settings and inputs to your forecast before clicking Submit.

Congratulations! You should now see your forecast inside of Trovata! Much like Reports, Forecasts are laid out with a visualization component on the top side of the screen, and a table of all your data down below. In order to override or update the forecast or actual data, you will need to visit the underlying data stream and select edit to append any values.

The Forecast line items will each have a letter next to them indicating what type of line item is being represented. The different letters stand for the following:

S = Stream

If you click into this line item you will be directed to the data stream. Here you can append the forecasted and actual values so long as you have the required access to do so.

G = Group

Here you are able to bucket a number of data streams under one group and see the total amount for these streams rolled up. You can click into the group at any stage to drill down into the individual data streams and see their individual amounts.

F = Forecast

Here you are able to see an overall forecast view. You can click into the forecast to further see the line items and detail that make up this forecast.

Each line item will include:

Total Cashflow: The sum of the total line items above.

Actual Cash Balance: The sum of the closing account balances for that day.

Daily Cadence Forecasted Cash Balance: Forecasted cash balance for current day is equal to the previous day actual cash balance + current day forecasted total cash flow.

Weekly Cadence Forecasted Cash Balance: Forecasted cash balance for current week is equal to the previous week actual cash balance (Actual CB shows the balance as of the close on Friday) + current week forecasted total cash flow.

Monthly Cadence Forecasted Cash Balance: Forecasted cash balance for current month is equal to the previous month actual cash balance (Actual CB shows the balance as of the close of the last day of the month) + current month forecasted total cash flow.

Quarterly Cadence Forecasted Cash Balance: Forecasted cash balance for current quarter is equal to the previous quarter actual cash balance (Actual CB shows the balance as of the close of the last day of the quarter) + current quarter forecasted total cash flow.

Forecast-Roll Ups can be used to consolidate forecasts across multiple Entities into a consolidated forecast view.

To create a Forecast-Roll Up, click New Forecast and complete Step 1 Basic Info, followed by choosing the accounts associated with all the Forecasts being consolidated.

Click Add Sources → Select source(s).

Click the Forecasts tab and select the Forecasts from the list that are to be consolidated into a single Forecast-Roll Up.

Scenario modeling a forecast in any software is just as much art as it is science.

The #1 question to always ask is: “What question are we trying to answer by modeling this scenario?”

The tools in Trovata to create and manage various scenarios will be covered in detail below:

To being modeling additional scenarios in Trovata, it’s suggested to begin by duplicating a “base case” forecast scenario. Usually, these base case scenarios make conservative estimates for all inflows and outflows.

In the “base case” forecast, navigate to the top right corner, select the 3 dots, and then click Duplicate.

Input a New Forecast Name for this new forecast and click Save.

The forecast has now been duplicated with the exact same values, forecast assumptions, and settings. At this point, this new forecast is exactly the same as the original.

Now, tools like Appending Values or Factors can be used to manuipulate the forecast data and create a new scenario.

It is important to note that once you create a data stream you can always make edits. To make changes to the existing setup, navigate to the top right corner, select the 3 dots, and then select the ‘Append Values’ option.

Within the 'Append Values' menu, there are various actions you can perform:

Factors can be applied to individual data streams or entire forecasts.

There are two types of factors.

Forecast maintenance requires diligence.

The built-in features of Trovata make it easy to report actuals against forecast. However there are areas where human intervention is required, like editing streams, updating accounts, and adjusting assumptions.

A common reasons to edit a streams is when there is a new type of cash flow.

Occasionally account strutures change. In a forecast, this means that cash flows associated with certain accounts may not occur anymore or may occur in a different account.

As the business evolves, so do the assumptions within the forecast. Adjusting assumptions may mean appending values, applying new growth factors, or rebuilding the stream structure in a forecast altogether.

Forecast reconciliation is the exercise of comparing what was forecast vs. what actually occurred, then adjusting the assumptions for future forecasts accordingly.

Within Trovata, current period Actuals can be investigate by directly clicking on the Actuals value in the forecast.

From there, adjust future forecasts as necessary by appending the stream values using the Update Data setting of the stream.

As previously mentioned, once you create a forecast, it's very simple to make changes to the existing setup. Simply navigate to the top right corner and select the 3 dots. More detail on the various actions you can perform can be found below:

You can change general details like the name of the forecast, default date range, currency, cadence, and rounding figures.
​

Here you can add or remove accounts to your forecast.

Here you can add or remove streams or forecasts to/from your forecast.

Here you can create Groups and bucket your data streams. Example: Receipts & Disbursements or Inflows & Outflows.

Here you can add factors to individual data streams or entire forecasts (covered previously).

Duplicate a forecast for scenario analysis or simply just to use a previous forecast as a baseline.

Ability to export the forecast to PDF.

An audit trail to track if and when changes are made to the forecasts.

You will be prompted with an “Are you sure?" warning message "Deleting this forecast is permanent and cannot be undone". Click ‘Yes’ to proceed or ‘No’ to cancel.

You can change the cadence view of your forecast at any time, either through settings or an ‘on the fly’ change. Daily, Weekly, Monthly & quarterly are your current selection options.

Data Stream and Stream Group ordering can be changed by toggling the pencil icon.

Here you have the ability to show or hide either actuals or variances in your view.

If you want to export this data into a spreadsheet, you can use Grid Mode via the waffle icon and click the top left most cell to highlight everything for a copy/paste of the data.

Cash forecasting is important because it provides visibility into all the variables that could affect future liquidity.

The most important aspects of cash forecasting are the assumptions used to create the forecast.

We designed forecasting in Trovata to give our customers an Excel alternative to store and manage those assumptions, as well as provide automated reporting of actual cash flows for cleaner variance reporting.

After creating the forecast and monitoring for variances, you will learn new insights about your cash flows.

Also, by centralizing your Forecasts inside Trovata, you can give your team access to own and control their data, all while avoiding Excel errors, file corruption, and manually reporting cash flow actuals.