The difficulties individuals face when processing numbers are further complicated by the nature of interest rates. Interpretation of interest rates requires not only the recognition of prices, yet also calculations of the 'value' of these yields; actions difficult for many people. (Bogomolova, 2009) Responding to these concerns, UK financial regulators, including the Financial Services Authority (FSA) and the Office of Fair Trading (OFT), have examined the forms of advertising and clarity of information provided on a wide range of financial services products. (Waite, 2006)
To summarise, price clustering is recognised in many different markets and a diversity of factors are advocated by way of explanation. (Gambacorta, 2008) A consistent theme within these explanations is the motivation of price setters. Often price clustering has been viewed as a phenomenon largely caused by exogenous factors or as a benign activity to assist customers and firms when assessing and setting prices. (Johnson, 2005) Alternatively, price clustering can be seen as an action undertaken by price setters to maximise returns from consumers with less ability to recall and process number information. From this review a model is developed which emphasises this behavioural interpretation of price clustering. (White, 2004)
Deposit interest rate observations at integers, half points and at quarters occur in 40-60% of all observations. For all quantities deposited there are lower levels of integer, half and quarter point observations as the value of the deposit increases. For mortgages, integers, half points, quarters and round numbers do not frequently occur and are recorded in less than 20% of observations. (Ashton, 2008) The proportion of firms setting interest rates using integers and half points also varies substantially between deposits and mortgages. Over 75% of deposit issuing firms set interest rates using both integers and half points compared with 35% of all mortgage issuing firms.
More deposit interest rate observations are located in the 50 basis points range above integers than below integers for all quantities invested. (Cameron, 2006) In accordance with the model predictions this is evidence of naïve truncation at the integer level. (Harrison, 2000) Similarly, more deposit interest rate observations are located in the 25 basis points range above integers. When a 5 basis points range is examined in relation to integers, half points and mid-points, the deposit rate observations have a very strong tendency to be above rather than below the reference points. These findings are supportive of the hypothesis that some depositors engage in naïve assessment of numbers. (Heffernan, 2002)
More mortgage interest rate observations are distributed in the 50 basis points range below integers (60% of total, than are recorded above integers. (Hall, 2009) This supports the hypothesis of naïve truncation at the integer level. A higher proportion of mortgage observations are also recorded in the 25 basis points range below integers and half points (47%), rather than above these points (39%). Only a small proportion of interest rate observations are recorded 5 basis points above and below integers, half points, quarters and round numbers; a proportion which increases ...