The Black Box of Mutual Fund Fees

This paper re-examines the determinants of mutual fund fees paid by mutual fund shareholders for management costs and other expenses. There are two novelties with respect to previous studies. First, each type of fee is explained separately. Second, the paper employs a new dataset consisting of Spanish mutual funds, making it the second paper to study mutual fund fees outside the US market. Furthermore, the Spanish market has three interesting characteristics: (i) both distribution and management are highly dominated by banks and savings banks, which points towards potential conflicts of interest; (ii) Spanish mutual fund law imposes caps on all types of fees; and (iii) Spain ranks first in terms of average mutual fund fees among similar countries. We find significant differences in mutual fund fees not explained by the fund’s investment objective. For instance, management companies owned by banks and savings banks charge higher management fees and redemption fees to nonguaranteed funds. Also, investors in older non-guaranteed funds and non-guaranteed funds with a lower average investment are more likely to end up paying higher management fees. Moreover, there is clear evidence that some mutual funds enjoy better conditions from custodial institutions than others. In contrast to evidence from the US market, larger funds are not associated with lower fees, but with higher custody fees for guaranteed funds and higher redemption fees for both types of funds. Finally, fee-setting by mutual funds is not related to fund before-fee performance.

Music-Theoretic Estimation of Chords and Keys from Audio

This paper proposes a new method for local key and chord estimation from audio signals. This method relies primarily on principles from music theory, and does not require any training on a corpus of labelled audio files. A harmonic content of the musical piece is first extracted by computing a set of chroma vectors. A set of chord/key pairs is selected for every frame by correlation with fixed chord and key templates. An acyclic harmonic graph is constructed with these pairs as vertices, using a musical distance to weigh its edges. Finally, the sequences of chords and keys are obtained by finding the best path in the graph using dynamic programming. The proposed method allows a mutual chord and key estimation. It is evaluated on a corpus composed of Beatles songs for both the local key estimation and chord recognition tasks, as well as a larger corpus composed of songs taken from the Billboard dataset.