diff --git a/forecast_client/README.md b/forecast_client/README.md
index 539c156..6282ff0 100644
--- a/forecast_client/README.md
+++ b/forecast_client/README.md
@@ -7,7 +7,7 @@ pip install -r requirements.txt
 ``
 
 ## Setup
-Fill-in the `conf.ini` file with the 
+Fill-in the `conf.ini` file with the
 appropriate endpoints and DB credentials.
 
 ## Usage
@@ -23,7 +23,7 @@ UDRs, CDRs and bills based on the forecasting models.
 Read the next section for more details on generating
 forecasts.
 - Use `cleanup` to delete all forecast records
-generated by the `--target` model you specify. You can optionally 
+generated by the `--target` model you specify. You can optionally
 delete `--all` rules associated
 with the model.
 
@@ -38,7 +38,7 @@ rules. This is to ensure it gets checked first. *This will
 not affect real records as long as the next guideline
 is also followed.*
 3. The rule must check the records that fire it for
-the name of its forecasting model. Records created 
+the name of its forecasting model. Records created
 by the forecasting engine are tagged in the account
 and data fields. You can check either one.
 
@@ -47,13 +47,9 @@ There are 3 types of forecast:
 1. `single` account forecast: A usage and cost estimate
 for a single `--account`. You will need to provide the
 account name, the pricing `--model` to be used and the
-forecast `--length` in days. Usage records are generated
-using the ARIMA forecasting model.
+forecast `--length` in days. Usage, charge and billing
+records are generated automatically.
 2. `global` forecast: As above, but provides a
 global estimate that aggregates data from all accounts.
 3. `pattern` forecast: Generates a forecast for overall
-usage for every account, also taking into account
-usage and activity fluctuation. It generates an activity
-pattern using the ARIMA model and then a per-account 
-estimate also using the ARIMA model, hence the name.
- 
+usage for every account.