piraeus bank group

"Over" and "Under" Valued Financial Institutions: Evidence from a “Fair-Value” P/B Econometric Model

Part of our mandate in the formulation of the investment strategy of the Bank is to be able to identify and evaluate investment opportunities (either in the Fixed Income or in Equity space) within the financial institutions’ universe.

The aim of the present study is to introduce a process that allows us to evaluate the relative “overvalue” or “undervalue” of the shortlisted financial institutions based on the distance between their market-based price-to-book ratios (hereafter P/B) and our estimated “fair-value” P/B.

With the help of three categories of variables that describe asset and asset quality, capital adequacy and liquidity, and efficiency and profitability, we come to the following conclusions:

  • Estimated coefficients are statistically significant (with the exception of the net interest margin) and have in most cases the economically expected sign.
  • The model penalizes size given that the log(Assets) enter the model with a negative sign as well as financial institutions with low credit quality (negative coefficient on Problem Loans / Gross Loans" (" PL2GL) and rewards high level of provisioning (positive coefficient on Liquid Banking Assets / Tangible Banking Assets (LLP2GL)).
  • High levels of profitability (ROA) and the efficiency (low Cost to Income (C2I)) are also having a positive impact on P/Bs.
  • High levels of TIER1 also push fair values of P/Bs higher but not the Shareholders’ Equity / Total Assets (Seq2TA).
  • Finally, liquidity variables have a somewhat contradictory impact given that high levels of gross loans to deposits (GL2TD) are penalized but so too high levels of liquid assets to tangible banking assets (LBA2TBA). A possible explanation for this finding is that liquid assets improve liquidity but harm profitability given their low yielding capacity.
  • As a robustness check exercise, we split the period into two subperiods. The significance of the variables changes keeping some commonalities to the aggregate model, mainly with respect to the ability of the model to explain the variation of the P/B. This finding is of great importance, considering the international panel setting.