Published March 2014

Market Statistics - How accurate are they and are they reliable?

By The Capital Corporation

The first question we often get from bankers is - What are banks selling for today? 


Last year we prepared our first Market Statistic Newsletter reporting on our transactions.  This newsletter reports on our 2013 statistics and compares them to our 2012 statistics.




Bankers and advisors alike have always looked at market statistics on multiples of book as a method to determine value.  Clearly a multiple of book is not the only way to determine value, and in our opinion it is overused.  Nevertheless, we all look at the multiple of book value when we see a transaction reported.


Market statistics are simply an average of the transactions included in the pool of sales that is used when reporting the average valuations. For example, the statistics that we are providing are transactions that we were directly involved in during the past year.  Therefore, it may be a smaller sample than many reported, but we also know that it is accurate and the limited geography of the transactions is relevant to you.  The point is that if you don’t know what the transactions are in the pool of statistics you don’t know if the statistics are accurate and relevant.


From 2003 to 2007 there were very few “problem” institutions.  Valuation differences occurred primarily based upon location, size and earnings.  Today, however, the condition of the bank is a significant factor in determining value in many

transactions. We see transactions occurring everywhere from banks that are close to failing to banks that have record earnings, therefore it is very difficult to gauge the average true market value of banks.


For example, if a pool of transactions includes the sale of several problem institutions at below book and several clean organizations selling above book value the average multiple of book may be close to book – even if none of the transactions were actually near book value.  The range of multiples in the pool can have a significant impact on the “average” multiple.




Another factor effecting reported multiples is the accuracy of the individual transactions included in the pools. 


We don’t report pricing on our transactions, but over the years many of our transactions have been reported by others and several of them have not been accurate.  The way excess capital is handled at closing can impact reported multiples.  Often the excess capital is paid out as a dividend to the shareholders immediately prior to closing reducing the capital.  The price is then paid on the lower capital number.  But if a price is reported and you look at the capital from the most recent call report the reported price looks lower than the actual price when you calculate the excess capital dividend.

Another example of incorrectly reported transactions is where the reported price does not reflect future payments such as non-compete and/or consulting agreements.  Again, it is easy to see how a payment that is likely reflected in a separate document (non-compete agreement) can be missed by the reporting entity.  We have had deals in the past couple of years where the non-compete and/or consulting payments was a substantial portion of the premium paid. 




Escrows, earn-outs and the transfer of loans, OREO and charged off assets from the bank to the sellers as part of a transaction can also have a significant impact on a reported multiple compared to the actual multiple realized by the sellers.


Although debt, Trust Preferred Securities and TARP are at the holding company level they can impact how multiples are reported. As explained above, it is easy and fairly common for individual transactions to be incorrectly reported in publications.  


Below are some statistics from transactions that we have been directly involved with during the past year.  We know the multiples are accurate.  While the parties in most of our transactions are disclosed it is our policy to not release pricing on specific transactions. For that reason everything provided below is in summary format.




In the past year we have been directly involved in 14 whole bank transactions that were closed or reported in 2013.  Some of 2012’s transactions are included in the 2013’s statistics because we include announced deals where the final purchase price is set.  So several of our deals that were announced and included in the 2012 statistics were closed in 2013 and are included in this year’s newsletter. 


All of the banks in our analysis were located in Kansas, Missouri and Nebraska.  They range in size from approximately $15,000,000 in assets to just over $100,000,000 in assets. One of the banks would be considered a problem institution. The pricing is based upon the amount actually paid by the buyer to the seller; not what was originally offered. The average size was approximately $52,000,000.


The average price to capital (based on 8% capital) for the whole bank transactions was 1.15 times capital.  The range was from 0.84 times capital up to 1.46 times capital.  The median multiple of capital was 1.20.  As you can see the median and the average were similar, but there was a wide range of multiples.  For reference, the average price to total capital was 1.13, very similar to the multiple of 8% capital.


The earnings multiple in our transactions was 35.71 for those banks that had earnings. We had three banks that reported losses in the twelve months prior to sale that were excluded from this analysis.   Additionally, if you remove the banks which had an earnings multiple higher than 30 times earnings the earnings multiple drops to 19.47. 


Banks that are selling for over 30 times earnings are clearly not selling based upon the earnings of the institution.   Excluding the loss and high earnings multiples, the range was from 11.75 to 27.23 times earnings.  The median multiple of earnings was 18.81.  As you can see, the range is fairly large but the median and average are similar when removing the loss and very high multiples.


When we look at values and multiples we also look at the premium as a percentage of the deposits.  When evaluating a whole bank transaction as a premium on deposits, you take the purchase price less the tangible capital to determine the premium and then divide that premium by the total (or core) deposits.  Basically, the purchase price is the premium on deposits plus the tangible capital.

Therefore, if you analyze the whole bank transactions as a premium on deposits, the average premium on deposits in those transactions was 2.62%.  We had one transaction that closed below book value which was removed from this analysis.  The range on these transactions was zero to 4.57%.  The median was 2.12% excluding the negative transaction.


Below is the 2012 chart for comparison to the new 2013 chart.



As you can see from a comparison of the two charts, prices on our transactions dropped from 2012 to 2013.  We believe that this is based more upon the location of our transactions this year as opposed to the market as a whole.  The banks in 2012 averaged $50,600,000 in assets, and the banks in 2013 averaged $52,000,000 in assets.  Clearly, this size range of banks seem to be the most consistent sellers in our markets. 

What we are currently seeing, both in the reported transactions above and the other transactions we are working on that have not yet been announced, is that many of today’s sellers are banks with limited growth opportunities.  Many of these institutions are also smaller which can reduce the number of interested buyers.  These factors are driving pricing more than we have ever seen in the past.   

Historically, any clean bank with earnings would generate substantial interest from buyers and thereby drive pricing.  Today, however, we see buyers be much more selective in their interests.  We still have many buyers for banks in or near strong markets, whereas for banks in markets with limited loan demand and growth opportunities the number of buyers is fewer.  This supply and demand directly impacts pricing.  We believe that we will see pricing remain low for banks with limited opportunities and pricing remain strong for banks with significant loan and growth opportunities. 



As you can see below, other reporting agencies offered different but somewhat similar, multiples.  Keep in mind that the following multiples are often on total capital, not 8% capital.  SNL Securities reported a price to book of 1.25 on transactions in the Midwest.  Thompson Reuters reported a price to book of 1.40 on transactions involving banks under $100,000,000 in the Midwest.  They reported a price to book of 1.02 on transactions involving banks of all asset sizes in the Midwest.


As you can see different reporting services as well as different geographic and size parameters will influence the resulting multiples.  In most cases the range of multiples is very wide from top to bottom which means that the reported average or median price may or may not have any applicability to your bank value. 


Thank you for letting us share our thoughts on the banking industry with you.   If you have any questions or comments, please feel free to give us a call.