I am starting to look at a more formal way to integrate Asset-Liability Management (ALM) into our investment process. We are a investment advisory firm that performs a scaled down version of what a typical “wealth management” firm would offer. Our clients are individuals. Right now my firm uses an Asset Only (AO) approach that indirectly models the quasi-liabilities (client goals) through the return requirement. I believe we can do better with ALM. I am a charterholder and have exposure to the topic through the CFA program. While this was helpful it is mostly academic and the mechanics of implementation aren’t explored. Does anybody have any experience in this area? Can you lead me to any resources to help me out?? Edit: Here is an example of what I am talking about http://www.cfainstitute.org/learning/products/multimedia/Pages/42913.aspx?WPID=Topic_List_Tabbed&PageName=All
Hey mwvt, If I understand you correctly, you’re looking for a way to implement a way to measure to match liabilities to assets for personal portfolios? So cash would be matched to the risk of a job loss or big unexpected expense. Bonds would be for inflation risk. And stocks would be for appreciation and aspiration? An article you may wish to check out is Beyond Markowitz, A Comprehensive Wealth Allocation Framework. Here the investor objectives and constraints are looked at through 3 frames. If you wish I could send you an overview in email. Otherwise the full article is available on the CFA website for $3. Nice web-cast in any case.
Since when do you use bonds for inflation risk?
QuantJock_MBA Wrote: ------------------------------------------------------- > Hey mwvt, > > If I understand you correctly, you’re looking for > a way to implement a way to measure to match > liabilities to assets for personal portfolios? So > cash would be matched to the risk of a job loss or > big unexpected expense. Bonds would be for > inflation risk. And stocks would be for > appreciation and aspiration? > > An article you may wish to check out is Beyond > Markowitz, A Comprehensive Wealth Allocation > Framework. > > Here the investor objectives and constraints are > looked at through 3 frames. If you wish I could > send you an overview in email. Otherwise the full > article is available on the CFA website for $3. > > Nice web-cast in any case. What I am looking for is an approach similar to what pension funds use. Basically I would be looking to duration match asset and liabilities to an extent. So risk would no longer be defined as standard deviation but a shortfall risk of funding quasi-liabilities. The idea is covered in L3. I will check out the paper quant. Thanks.
Bebop and Rocksteady Wrote: ------------------------------------------------------- > Since when do you use bonds for inflation risk? TIPS?
Chuckrox8 Wrote: ------------------------------------------------------- > Bebop and Rocksteady Wrote: > -------------------------------------------------- > ----- > > Since when do you use bonds for inflation risk? > > TIPS? I meant the rate of return is similar to the rate of inflation in most cases. Commodities would be more ideal for inflation risk.
Hey mwvt9. This is an interesting topic and something that I have thought about in the past. I think there are some barriers which make ALM difficult on an individual level; primarily account size, turnover due to rebalancing, and difficulty in measuring duration for many of the assets and future liabilities that exist for individual investors. The duration matching characteristics of bonds make them excellent for matching statistically sound duration estimates for large institutional funds (ie pensions). This becomes more difficult when we include equity and human capital onto the asset side of the equation. Would you be including all tangible and intangible assets into the ALM models? If so, you would obviously need to make some pretty big assumptions about income and maturity for equity and intangible assets. I still think you could come up with a somewhat reasonable estimation of duration based upon well-grounded assumptions. It would be interesting to see the result of this type of ALM approach and compare it to the general asset only MPT models that are so prevalent in the academic world. There has to be room for improvement by using ALM and shortfall risk compared to the classic approach which classifies risk as nothing more than a willingness/ability to accept normal market volatility…
I found the ALM stuff in the L3 curriculum to be quite interesting, and it certainly comes across as more appropriate than the traditional asset-only approach. But like McLeod81 says, the characteristics of individuals vs those of institutions probably makes an ALM approach tough to implement. It might be possible to use some of the tools or concepts for very large clients (asset-wise) I guess.
mwvt, Here is an article that seems relevant as it defines an ALM framework for individuals (disclaimer, I haven’t read the arcticle): http://www.edhec-risk.com/edhec_publications/all_publications/RISKReview.2009-10-07.3553/attachments/EDHEC_publication_ALM%20in%20PWM.pdf is that what you are looking for?
maratikus Wrote: ------------------------------------------------------- > mwvt, > > Here is an article that seems relevant as it > defines an ALM framework for individuals > (disclaimer, I haven’t read the arcticle): > > http://www.edhec-risk.com/edhec_publications/all_p > ublications/RISKReview.2009-10-07.3553/attachments > /EDHEC_publication_ALM%20in%20PWM.pdf > > is that what you are looking for? This was the best article that my research could turn up. I have skimmed it at this point and plan on looking over it in more depth. Maybe you can help me with the math, eh?
McLeod81 Wrote: ------------------------------------------------------- > Hey mwvt9. This is an interesting topic and > something that I have thought about in the past. > I think there are some barriers which make ALM > difficult on an individual level; primarily > account size, turnover due to rebalancing, and > difficulty in measuring duration for many of the > assets and future liabilities that exist for > individual investors. > > The duration matching characteristics of bonds > make them excellent for matching statistically > sound duration estimates for large institutional > funds (ie pensions). This becomes more difficult > when we include equity and human capital onto the > asset side of the equation. Hey McLeod how are you? I agree with everything you have said here. > > Would you be including all tangible and intangible > assets into the ALM models? I would likely include human capital and say social security as illiquid, untradable assets. Quasi-liabilities (client goals) could be treated like zeros to come up with a rough duration. > > If so, you would obviously need to make some > pretty big assumptions about income and maturity > for equity and intangible assets. I still think > you could come up with a somewhat reasonable > estimation of duration based upon well-grounded > assumptions. Our standard planning has to make some big assumptions anyway. I guess my thought is to redefine risk first - from standard deviation in the AO approach to shortfall risk in ALM. One that is done we could show a basically immunized portfolio taking all shortfall risk out of the picture. Of course this will be the most expensive option. We would then educate clients on the trade-off of moving to the less expensive, but higher shortfall risk portfolios (which would likely include equities). In this manner the client goals become the driver for portfolio decisions (which is how we do it now anyway) and we move away from did I beat the SP 500 this week. So ultimately it wouldn’t be as technical as a pension ALM because, let face it, we don’t have legal liabilities here and we can change them at any time (Okay, I won’t buy the yacht), but it could still be an improvement to AO which doesn’t even recognize the liabilities. Also, for my firm which does financial planning too it gels nicely with the managing the entire balance sheet. > > It would be interesting to see the result of this > type of ALM approach and compare it to the general > asset only MPT models that are so prevalent in the > academic world. > > There has to be room for improvement by using ALM > and shortfall risk compared to the classic > approach which classifies risk as nothing more > than a willingness/ability to accept normal market > volatility…
pimpco has some articles on ALM on their website that you can search for
Thanks cubsfish. I’ll check it out.
mwvt9 Wrote: > This was the best article that my research could > turn up. I have skimmed it at this point and plan > on looking over it in more depth. > > Maybe you can help me with the math, eh? No problem. Shoot me an email if you have specific questions.