Modeliranje likvidnostne premije na podjetniških obveznicah : delo diplomskega seminarja

Perme, Jan (author), Bernik, Janez (mentor)

Likvidnostna premija je ena od lasnosti obveznice, ki je ne moremo opazovati na trgu. Da bi jo izračunali, tako konstruiramo model, ki temelji na tri stopenjskem postopku. V prvi fazi z regresijo modeliramo razpon ponudbe in povpraševanja, ki je kazalnik likvidnosti, kot funkcijo karakteristik obveznice. Tako dobimo nov kazalnik likvidnosti, neodvisen od karakteristik obveznice, ki ga poimenujemo $RBAS$. V drugem koraku modeliramo kreditni razpon kot funkcijo karakteristik obveznice, ter $RBAS$-a. V tretji fazi ekstrapoliramo $RBAS$ v regresijskem modelu v 0, ter tako dobimo kreditni razpon enake, a popolnoma likvidne obveznice. Od kreditnega razpona obveznice odštejemo kreditni razpon popolnoma likvidne obveznice in dobimo likvidnostno premijo. Nato naredimo nov regresijski model, ki likvidnost transakcije obravnava v odvisnosti od likvidnosti obveznice, velikosti transakcije, dnevnega donosa, časa ter tega kdo je začel nakup oziroma prodajo. Model uporabimo v praksi in ugotovimo, da je razpon ponudbe in povpraševanja najbolj odvisen od trajanja obveznice in nominalne vrednosti. Prav tako ugotovimo, da je kreditni razpon obveznice močno odvisen od nominalne vrednosti in trajanja.

• Language: Slovenian
• Keywords: likvidnostna premija, razpon ponudbe in povpraševanja, kreditni razpon, obveznice
• Work type: Final seminar paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FMF - Faculty of Mathematics and Physics
• Year: 2018
• PID: 20.500.12556/RUL-103257
• UDC: 519.8
• COBISS.SI-ID: 18435929
• Publication date in RUL: 15.09.2018

Secondary language

• Language: English
• Title: Modeling the liquidity premium on corporate bonds

Abstract

Liquidity premium is one of characteristics that we cannot observe on the market. We construct a model that is based on a three stage recipe to compute it. In the first stage we use regression to model bid-ask spread, a liquidity proxy, as a function of bond characteristics. As a result, we get a liquidity proxy independent of bonds characteristics that we name $RBAS$. In the second stage we model credit spread as a function of bond characteristics and $RBAS$. In the third stage we extrapolate $RBAS$ in the regression model to 0 wich gives us credit spread of an identical but perfectly liquid bond. We then deduct this prefectly liquid credit spread away from the bonds credit spread and get a liquidity premium. We then construct another regression model that models liquidity of a transaction as a function of bond liquidity, size of transaction, daily yield, time and who initiated the transaction. We use this model in practice and observe that the bid-ask spread is highly dependent on bond duration and nominal value. We also observe that the credit spread depends on nominal value and duration of a bond.

• Keywords: Liquidity premium, bid-ask spread, credit spread, bonds