One consequence of the current rise in credit delinquencies is that a growing percentage of borrowers’ risk profiles are changing quickly. As a result, if lenders aren’t getting updated account management scores frequently enough, they run the risk of making decisions with stale information on a significant portion of their borrowers’ accounts.

Lenders are asking: How often should I refresh the FICO® scores of my existing customers? How many of my customers will have a significant change in their score in the short term?  How much will their scores change, and over what period of time?

Fair Isaac set out to find the answers. We examined recent “score migration” trends to determine what percentage of consumers’ scores change and the rate of change over different periods of time.

Key findings

For this study, we looked at the extent of score migration in a matched sample of more than 100,000 active accounts in a prime bankcard portfolio over three-month, six-month and nine-month time periods. We performed the same study twice over two years and the results were consistent.

Here’s what we found:

• A significant percentage of scores migrated up or down more than 20 points over one, two or three quarters—from 25% in the first quarter to as much as 42% by the end of the third. Even though the majority of individuals’ scores remain relatively constant, a significant percentage of the population will have score changes that are large enough to be problematic and warrant account management attention. (And results may be even greater on a nonprime portfolio.)  A large change on even a minority of accounts in a portfolio may have a dramatic impact on portfolio risk and profitability.
  
• The overall pattern of score migration remains highly consistent. At any given point in time, approximately 25% of the population will have had a shift in their score of more than 20 points over the last three months.
  
• Higher scores tend to remain more stable over time. Scores in the lower range are more likely to fluctuate, whether upward or downward. (A change in a low score does not necessarily mean that account is destined to continually decline.)
  
• The most current score for a given account is the most predictive. Tracking the trend in that account’s scores over time is less useful as a predictive tool.

We also examined a separate sample to evaluate how much scores change from month to month, and found:

• A significant percentage of scores migrated up or down more than 20 points from month to month.  Multiple one-month snapshots reinforced the findings of the quarterly study, with approximately 17% of scores migrating up or down more than 20 points each month.

The findings suggest that refreshing FICO® scores of existing accounts on a quarterly basis at a minimum, and preferably monthly, will help lenders make more informed decisions on account treatment. Leveraging fresher, more accurate scores, lenders can not only target potential problems, but also identify candidates for more positive treatment and upselling.

How much do scores change over time?

To assess score volatility, our study examined how scores migrated over one, two and three quarters. For the majority of accounts, scores did not change more than 20 points upward or downward in the prior quarter.

However, 25%-27% of scores did change by more than 20 points over the preceding quarter, 35%-37% changed over the prior two quarters, and up to 44% changed by more than 20 points over the preceding three quarters. We further found that the longer the time period since score updates or refreshes, the more likely migration was to have occurred.

Most accounts’ FICO® scores stayed relatively stable over one, two and three quarters. However, 25% changed by more than 20 points in one quarter and up to 42% over three—significant enough to alter the balance of risk. The findings are consistent from one year to the next. (*Positive score difference equates to score increase over time.) 

Which accounts’ scores are more likely to migrate?

Fair Isaac examined score ranges to see whether certain subpopulations are more prone to migration than others.

In Figure 2, we see that higher scores are more likely to remain stable. For example, 71% of records with a score of 700-749 had scored within 20 points of their score three months prior. Higher scores (750+) are even more likely to remain stable in a three-month period, with 79% of these records staying within 20 points of their three-month-old score.

Lower scores, on the other hand, are more likely to fluctuate. In the under-550 segment, 47% of the scores—less than half—had migrated less than 20 points in the prior quarter. Of accounts with scores under 650 that had migrated more than 20 points in the prior three-month period, the majority tended to move downward.

Higher scores remain more stable. Lower scores exhibit more fluctuation, with most of those that change by more than 20 points moving downward.

Higher-scoring accounts will likely include the majority of a lender’s best customers. Knowing which high scoring accounts remain stable will guide lenders in taking appropriate action to retain their best customers. Conversely, lenders will want to know which of their high-scoring accounts have seen a significant drop in score, even if this is only a small percentage of total accounts.

How much more predictive is a fresher score?

We conducted a score cutoff sensitivity analysis to see to what extent a current score outperforms an older one as a predictor of risk. The results, as shown in Figure 3 for the first snapshot of accounts studied, illustrate the value of the fresher score.

• The first column in the chart shows sample score cutoffs.
  
• The next two columns show the percentage of accounts falling above and below each cutoff.
  
• The next two columns show the percentages of accounts that migrated above or below the cutoff three months after the initial observation date.
  
• The last two columns illustrate the odds on the “swap-in” (migrated above) and “swap-out” (migrated below) groups.

The older scores exhibit larger swap sets—for example, at the 690 cutoff, 3.1% migrated above at eight months compared to 1.5% at three months. They also show greater differences between the actual risk of the swap sets. For example, after eight months at the 730 cutoff, the odds for the “migrated above” set are 75.7 to 1 versus 23.9 to 1 for the “migrated below” segment, compared with 58.3 to 1 versus 34.3 to 1 after three months.

The older scores clearly provide a less accurate assessment of risk. Lenders using the older scores are at greater risk of making suboptimal decisions on consumers whose scores have changed.

The more time between score updates, the greater the discrepancy between the “migrated above” and “migrated below” odds. When the scores are 6-8 months old, the swap-in (migrate above) odds are between three and four times the swap-out (migrate below) odds.

Putting fresh scores to work

Fair Isaac considers the best practice to be monthly score refreshes.

We see more lenders moving from quarterly to monthly refreshes. Lenders also employ trigger mechanisms for pulling fresh scores on particular accounts in certain situations: “push” triggers, instructing credit bureaus to supply a fresh score when something significant, such as a new delinquency, happens on a consumer’s credit file; and “pull” triggers that tell the lender to obtain a fresh score, for example, when the customer is seeking a line increase.

Smart lenders want to be sure they are not caught unaware of borrowing activity that may turn yesterday’s “good” customer into a future write-off.