Finding factors in fixed income

[Applause] [Music] [Applause] [Music] our next speaker is jean-marie dumas the global head of fixed income solutions at mundi asset management he'll be talking about their approach to factor investing within fixed income [Music] thank you for the introduction ray and so the topic uh we'll present to you today um is about facto investing in fixed income so we will try to see what it represents in the new investment world i would say um how we can use it for sovereignty investment and also for corporate bond investment where i would say the factor investing is probably more accurate and to finish i will show how we can combine factors to use a multi-factor technique or multi-factor strategy chloe the first important thing is to know what we think about when we are talking about factor investing compared to the traditional way of building an investment strategy and particularly in fixed income the challenge of the factory investing is to identify alternative factors the one which can change a little bit the way to build portfolio compared to the traditional way of splits in alpha and beta strategy the the factors are in fact so alternative beta that we define on an excellent basis and that we can after catch the trends of the market and also identify some behavioral bias in the in the fixed income market of course it started long time ago in equity investments but a lot of new developments came over the last years to adapt it to equity long short but also to fix income and credit and also to cross-asset that's where you found a lot of strategies named alternative risk premium strategy for example which are based on these alternative factors that the quant research from asset management firms but also from instrument investors can identify what is important to notice is that factor investing is not about drawdown management or about pure alpha catching it's about achieving a long term out performance by strategic allocation strategic allocation which is based on those alternative factors the most important thing and the best thing we can bring with a factor investing technique is the clarity and transparency of the portfolio construction and the fact is that with those factors we aim at enhancing diversification in the portfolio thanks to the low correlation of these alternative factors with the traditional active strategies especially in fixed income for example the key point is to be able to build a strategy which is lowly correlated to for example duration or credit beta that's that's important or if it's correlated the important is to know how this factor behave alongside with the the traditional beta in fact that's this transparency which is important to better adapt your portfolio of course there is heavy research to run before launching such strategies because you have to identify the factors to be sure that these factors are consistent over time that they are not just a recombination of traditional factors and because at the end it will uh it will be a sort of traditional strategy just repacked so we do not look at marketing factors but clearly at alternative ones and it is important thereafter if you combine your different intensity factors to know how they react with each other's that's a really important thing which will enable us to build a strategy where the tail risk is under control and where we know that the factors are significant and shows a statistical likelihood so factor investing is not a sort of magical pot you you have to control also your risk through traditional factors when you do so you have to that's what i call um the transparency effect that knowing how your alternative factors are correlated with duration with credit beta with liquidity which are the traditional factors and thanks to that you can have a well-structured portfolio so to go um into real world it is interesting to see how we can adapt it and use this technique for example server and bone investment because in fact traditional factors are very strong in sovereignty investors you know that you can represent the interest rate risk framework with the traditional duration curve um and butterfly effect so that's important to notice that with a traditional for example component principle and principal component analysis we can catch a large part of the volatility of the sovereign known market so it's not so easy to define alternative factors which are not just a recombination of the traditional one what we can see and what we what we did in the past is to identify alternative factors which work quite well um in the sovereign one world first of all the carry factor which is which can be described in two different factors the first one which is a time series carry factor and this one which is a cross section carry factor in fact this signal on the time series one is just um a way of catching the roll down effect on the curve so that is why uh the formula which is in this page eight shows that it's a duration adjusted difference between the yield on the 10 year 10 hour and the yield on the 9 year on 11 months tenor so that's just the effect on the rate curve of the natural roll down and you build your carry factor by choosing the bones which shows the most important rolled on over a global universe a global sovereign universe um on the uh on the cross section uh chi factor that's a little bit different that the whole curve of the of the phone that you that you play and the casino is just the 10-year government on yield minus the three months deposit rate and once again you rank uh your different kind signals and then you pick up the highest carry and short the lowest carry so these carry factors are in fact long short and in with that you you can show that you catch a special effect on the sovereign bond curve you can also play what we call momentum so the momentum factor is like what is done on ct investing meaning that you buy the bones which shows a good good recent performance and you short the lowest performers within your universe and so that's that's really non-discretionary you build your momentum factor by being long the performers and short the under performers of course what is interesting with the cta with the momentum factor is that is very lowly correlated to the carry factor and more than that to the duration traditional factor you see in the sovereign world that's what we have in this slide where the the the scatter plot shows that in fact the correlation between the return of the strategy which is here a combination of ct plus carry models and the german tenure yield is very low so thanks to that we can see that building such factor portfolio alternative factor portfolio in server and one world is not something really related to the duration and so by using such a strategy for example in overlay on top of a traditional bond phone brings performance but also an almost diversification so which is a good thing and that's the way you use it and it is very easy to implement because you can do this kind of factors you can replicate this kind of factors with the use of derivatives such as bond futures or swap so that's uh what we can say about sovereign bond alternative factors and of course this kind of candles can also be used in a multi-asset portfolio more interesting i would say than on sovereign bones where sometimes the alternative factors are less significant statistically as i said because the traditional factors are catching a lot of the volatility of the market and explain quite well the the return of of sovereign bonds we can use the factor investing technique in credit market it is a natural extension of what was done over the last 20 years on the equity market because on corporate incorporated market you are investing in instruments where the specific risk is important and so you can think that the alternative factors will help us to identify within the specific risk some commonalities that's what i will present right now in the corporate bone factor investing essentials first of all once again using this factor investing technique is here to describe a systematic approach a systematic approach which is diversifying existing strategies of course not all the factors are real in the 4.1 world very often you will see in literature people talking about the classical factors that we use in equity market which are value momentum quality or low risk for example and we we did an important operational research to see among those factors which can really explain the credit returns so when i talk about credit return i talk about only the performance coming from the spread variation so the credit is premier not about the interest rate effect and for the interest rate effect you can go back to the first section where i describe the automatic factors for sovereign bonds so really important to rely on factors which are consistent over time and which are statistically significant what we show from our research is that among the classical alternative factors the most significant one and the one which are showing a real risk premium are the value factor and the momentum factor for example the low risk is just a recombination of duration and rating so credit quality and tenor of maturity you you can see also the efficiency which is called size but which is in fact more a way to to be exposed to lower liquidity meaning less liquid bonds than really um behavioral bias in the market so roughly size is such a recombination of the liquidity factor which is a traditional one so let's go in the presentation of these two interesting alternative factors which are value and momentum what is important to notice is that factors can be used alternative vectors can be used to explain the market behavior is because now the crit market globally speaking in u.s but also in europe is quite deep and that's important because thanks to that we can have better data more accuracy in the price in the prices we contemplate and it brings in fact uh more robust models so that's that's the important thing and typically about what we uh we experienced over the last 10 years we have to notice that uh the markets in credit are quite different from a region to another and especially we believe in regional factors meaning that there is no global factors when we are talking about a value factor we are talking about a euro value factor or u.s dollar value factor and not on a global one and that's how we apply it also from sometimes from time to time and it's true also for momentum factors of course as we're talking about systematic strategies we have to take into account some operational constraints which are about market liquidity because applying a model in credit is a little bit different from applying it in equity and so you have to control your liquidity cost of trade here and on the cost of trading and you have also very often to apply a turnover control because for example if you let the momentum factor completely free um you will have a turnover of around 400 percent a year so meaning uh reshuffling the port for you four times which is of course too costly in terms of uh trading and you have also to look at market events because bonds are not like equity they have a birth they have death and so or maturity we have to deal with what happens on the primary market that's important to rebalance the factors from time to time and to benefit from also this excess premium you can catch on the primary market and you have also to manage the potential credit deteriorations that you can see through rating downgrades because very often um when we have when we launch a phone or strategy we can have constraints on rating it's not the same thing to to propose an investment grade strategy or crossover one or higher and there are also the some special events like corporate actions when the bonds are reached a little bit renegotiated by issuer with sometimes the new covenants and so on so how we build a multi-factor strategy so as i said on corporate wounds based on value factor and momentum factor what we call a value factor in fact is a way to estimate the difference between the [Music] interesting value of the bones and its relative value in the market it is important to notice that for us we are talking about this relative value we are not talking about interesting value it is not a merchant model adapted to corporate bonds and we do not use data which come from other market than credit market so we rely only on a market which has on data which has we which have the same frequency uh in terms of update as the market and so which are completely in line with the price evolution our value factor means that for each bond we want to compute a theoretical spread based on market data and then to compare it to the observed spread and so this distance gives you the value signal so the higher is the distance the more important the value is in the market and so that's how we build this value factor by picking in fact the bonds with the higher value signal so the spread is explained through an economical model by characteristic of the bones and market data such as [Music] spreads and price so here you will see in page 19 a description of this of this model and how in fact we compute it the main characteristic we use are the classical one which are analyzed by all the market practitioners which are very easily directing the sector to country risk the seniority as we have more and more subordinated bonds in the market face value materially and of course optionality because you have more and more collaborate bonds in the market as i said the difference between the obstacred and the theoretical spread coming from the model gives you the value potential of course from one to another there will be a different explanation of the value potential and from time to time it will be more linked to sectors or seniority or sometimes to register of course after the first step of estimation of the estimate we have to build a real factor so for that we keep the first top 100 bonds with the value potential and we skip out the first 10 bonds why because in the value factors there are very often there is very often what we call the value trap which are the bones which are in fact undervalued by the market but for good reasons and they have that's uh sometimes the leading indicator of a potential downgrade or even default or credit event that is why statistically to avoid to be trapped in this value trap we exclude the first the 10 first bones from the value factor so the value factor is in fact the 90 bonds resulting from that process we equally weight those bones to avoid to have a too important jump to default risk meaning that the the exposure to credit event is equally weighted on on the factor and we rebalance it on a monthly basis of course as i said we have to constantly monitor this factor because we have to be reactive if there are some corporate action or if there are some primary bonds coming in the market on the momentum factor it's quite straightforward like for the sovereign bond the momentum factor is the way to be long the performers and short under performers as we are talking here about corporate cash bonds we are not talking about a long short momentum factor we are talking about long only we will only buy the performers which are represented by the bonds which exhibit the best performer over the six months training excess return that's that's how we we do it um and of course we scaled it by spread duration to avoid to be biased towards the long maturity bonds what is interesting is that the momentum factor is really a complementary to the value factor here you will see the payoff of the two strategies on the left side this is the value strategy on the right side the momentum strategy in fact what we can see that the momentum factor in the natural market environment does not perform a lot it does not destroy value but it is not [Music] very performing but what is interesting is that if we go into uh bad performance territories on the credit market then the momentum strategy posts good performance so which means that if you apply in a multi-factor strategy a combination of the value factor and the momentum factor then you have an interesting payoff as a combined strategy which is the one we have on this slide so meaning that of course you will not uh earn money at any time but it means that in terms of bearish time in a bearish period in the credit market you will mitigate the loss you could have on the value strategy by the gain you will benefit from the momentum strategy so the momentum strategy is a quite good edge interesting one and not not very expensive for a portfolio that is why we inaudible strategy in fact we combine these two factors with a relative weight which which gives enough protection to the value vector because the value factor is interesting but has a big bias which is a high beta exposure so meaning that the value factor in fact is a sort of smart high beta strategy of course we could constrain it but we think it's more interesting to benefit from that to be transparent and to control it so that's that's how we can benefit from positive risk premium by the combination of alternative factors in corporate bond world of course once again it's not a magical point you as you can see you are long credit risk the only uh different is that you are not just long credit beta by using this alternative factor your reaction will be or the reaction of your portfolio will be a little bit different and so that is a good complementary to traditional bond investment long only one investment here i talked about long only alternative factors we could also design long short alternative factors but of course if we do so we cannot rely on the bond market because shorting a credit bond is very expensive and quite complicated if we do so we will use cds so credit default swaps but that's another story if we want to do that we have to recompute our factors based on cds market data not on the bond market that is very important that in at every time when we are talking about cone strategies and systemic one systematic one we need to be very careful on the the adequacy between the data we use and the strategy we build that is why once again uh at the time we do believe that corporate bond alternative factors on more than fixed income factor in the state must rely on data that comes from the fixed income market not from other markets yeah i think i went through the whole presentation um i'm sorry that this presentation and the think tank is done virtually and i would prefer to be with you and physically present i hope it will be done next year and i wish you all the best in this uh uncertain time but once again facing uncertainties we have to think more and to innovate more and that's what the factor investing can bring i would say for our traditional investments [Music] you

Channel: Momentum Investments

Views: 89

Rating: 5 out of 5

Keywords:

Id: OrWDkWcpuUA

Length: 27min 32sec (1652 seconds)

Published: Wed Sep 30 2020