Forthcoming changes to quoted interest rates data - 2019

By Dominic Tighe and Alister Ratcliffe

The Bank of England publishes statistics on interest rates offered to households by banks and building societies – ‘quoted rates’ – for a range of lending and deposit products. To make these statistics more representative of the advertised rates faced by households, from Q1 2019 these statistics will be based on a larger number of lenders and the products included will more closely match those bought by customers. Quoted rates on some new products will also be introduced. Further details on how these changes impact the data will be provided alongside the first publication.

Introduction

The Bank’s ‘quoted interest rates’ statistics show the average interest rates advertised by banks and building societies for household products. They cover mortgages, consumer credit and deposits. Rates from individual lenders are combined together using the Bank’s data on lending and deposit volumes to create weighted average series that are representative of the market.

In Q1 2019 changes will be introduced to ‘quoted rates’ so that the statistics continue to reflect the advertised products that most consumers experience as closely as possible. Changes include the expansion of the sample from which the data are collected, the addition of new series and the updating of the methodology used for calculating existing series.

1. Expansion of the sample

First, there will be an increase in the number of lenders in the sample. Interest rates from the whole market will be collected, compared with the current sample that covers at least 75% of the market for household lending and deposits and is consistent with the Effective Rates (ER) sample. This sample change means that smaller banks will be better represented in the data.

Expanding the sample will require a change to the weighting methods that are used to calculate average ‘quoted rates’. This is because many lenders in the new expanded sample are not included in the ER data collection, which has been mostly used to weight quoted rates data to date. The weighting method used for the new data will vary by product.

For mortgages, the FCA’s ‘Product Sales Data’ (PSD) will be used to calculate weights. The PSD captures details of all mortgage contracts at the point of origin. New weights will be based on banks’ lending volumes for the same product - for example the owner occupier 2 year fixed-rate 75% loan-to-value (LTV) mortgage - that a rate has been selected for. This will be a more accurate weight than the current method of using the closest approximation for a product from ER data. For example, the quoted interest rate on the product above is currently calculated using weights based on institution’s mortgage lending with a fixation period of up to 2 years.

For time deposits (deposits where notice is required to withdraw funds), the expansion of the sample means a variety of different data sources will need to be used to weight lenders’ rates together. For around 90% of the market the value of new time deposits from the ER data collection will continue to be used. The granularity of this data corresponds closely with the quoted rates products. For the remaining 10% of the market not covered by this collection, data on the stock of deposits collected on the Bank’s other statistical returns will be used for weighting.¹ For sight deposits (deposits available on demand), the stock of deposits collected on the Bank’s Balance Sheet (BE) return will be used to weight all lenders’ rates.

For credit card and personal loan quoted rates, the value of new lending from the Bank’s Lending to Individuals data will be used for weighting around 95% of the sample. The remainder of the sample will be weighted using data on the stock of lending from the Bank’s Balance Sheet (BE) returns. For overdrafts, the stock of ER data will be used to weight around 80% of the market, with the Balance Sheet (BE) data used to weight the remainder.

Using stock balances to weight a portion of the data for deposit and consumer credit series is not ideal because quoted interest rates data aim to capture rates available to new customers. However, the benefit of this approach is that it enables the inclusion of a wider range of lenders in the sample.

2. Changes to the definitions of existing series

To ensure that quoted rates data keeps up to date with changes in the economy, alterations will be made to the definitions on some series.

Quoted rate on ‘revert-to rate’ mortgages

The standard-variable rate mortgage (SVR) series will be renamed the ‘revert-to rate’ series. The aim of this change is to better represent the rate mortgages revert to at the end of their initial offer period. For many lenders this is the SVR, but for some there is a separate ‘revert to’ or ‘reversion’ rate. Including all reversion rates rather than just SVRs in the product selection reduces the rate on this series by an average of 29bps between 2013 and 2018.

Selection of quoted rate for the ‘representative’ credit card rate series

Credit cards that have higher usage and are available to most customers are selected for quoted rates. From Q1 2019, products that are labelled as ‘low rate’, ‘premium’ or are exclusive to one profession or demographic - which tend to have more competitive rates - will be explicitly excluded from the selection process for the representative credit card series. Product selection will be reviewed periodically.

Approach for selecting the most competitive mortgage product from each lender

House prices have increased since the last quoted rates review and the average mortgage size has tracked this increase. Product characteristics tend to vary with mortgages size, so increases in mortgage size need to be reflected in quoted rates statistics. In line with data from UK Finance and MLAR, the size of mortgage used for quoted rates product selection will be increased from £150k to £185k.

3. Move to monthly average, rather than end-month rates

Quoted rates will be published as monthly averages so that the definition more closely aligns with other statistics collected in the Bank. Currently, the Bank’s quoted rates show rates available at the end of the month.

The benefits of moving to an average across the month will be that mid-month product changes will be reflected in the data. In addition, if a product is withdrawn within the month, it will still be captured in the monthly data. This change will bring the quoted interest rates methodology more closely in line with that for effective interest rates, where monthly gross interest flows and average daily balances are used to calculate rates.

This change will have a small effect on quoted rates statistics. The average absolute difference between the 2 year fixed-rate 75% LTV mortgage calculated on a monthly average basis and the current series (calculated on an end-month basis) has been around 3bps since January 2015. The greatest difference between the two series is 12bps.

4. Additional series

As well as updating the methodology used to calculate existing series, new series will be published to provide a more complete view of the market (Table 1). For owner-occupier mortgages only one series is currently collected for 5-year fixed rate periods, so the additional series will give a better view of this growing section of the market. The buy-to-let sector was similarly under-represented. For credit cards the additional series will allow quoted interest rates on credit cards with particular incentives to be monitored.

Table 1: Quoted interest rates series (products highlighted are new)

¹ Form BE time deposit data will be used for weighting most of this 10%. For the less than 1% of the market not covered by this return, the stock of all deposits data collected by Form BT will be used.

For questions relating to this article please contact DSDRateTeam@bankofengland.co.uk or call +44 (0) 20 3461 6047.