My approach to the problem of hurdle rates and market valuation

[Lupo Lupus]

 

I wanted to share my approach for how to adjust the size of my equity portfolio, depending on the current valuation of stocks. I would be happy to hear your thoughts on it.

 

My approach is quite different from having a (constant) hurdle rate for new positions:

 

 

I do not have a hurdle rate at all. I simply invest in the most undervalued securities currently available. However, I vary my cash-equity allocation in response to the valuation of the “marginal” security (the currently best security to add or to delete from the portfolio). My through-the-cycle target cash-ratio (where cash includes safe bonds) is 15%. If the marginal security is very undervalued (likely to happen at time of market undervaluation) I can go down to a 0% cash-ratio. If it is overvalued, I can hold up to 30% in cash. The current target cash-ratio is made explicit at any point in time to anchor investment decisions. In addition, I also have a list of the two marginal securities, that is, the security I would buy if I increase my equity allocation, and the security to sell if I would lower my equity allocation.

 

The justification for this strategy as follows:

• It does not make sense to have a fixed hurdle rate. It implies that if the market is fully valued, I will not be invested but if it is several undervalued, I may have to leverage up big time.

• My approach permits keep being invested even if no undervalued security is available. I will still get in expectation the average return on equity, which is nothing to be sneezed at.

• The approach is effectively a (dynamic) asset allocation approach where an asset class is weighted according to its expected return. Note that this is not the same as market timing. Rather it is based on portfolio theory considerations (overvalued asset has lower return and higher risk and hence should receive a lower weight).

• By conditioning on the valuation of my portfolio, rather than the market I can take into account that undervalued securities relative to the market are not equally available. For example, during the dotcom bubble the market overall was overvalued, while value still had a decent implied performance.

• The conditioning is on the valuation of the marginal security rather than the average security in my portfolio as the impact of variations in the size of my equity allocation will depend on the marginal security.

 

[clutch]

• By conditioning on the valuation of my portfolio, rather than the market I can take into account that undervalued securities relative to the market are not equally available. For example, during the dotcom bubble the market overall was overvalued, while value still had a decent implied performance.

 

I didn't get this part - could you please clarify?

 

In general, I see a couple of limitations in this approach. The first big limitation is that it seems like this approach is less goal-oriented than the hurdle rate-based approach. In the hurdle rate-based approach, you set a specific and absolute target rate, which motivates you to find a very few best investment opportunities. I feel like this mind set makes you work really hard to find the best opportunities. The concept of "marginal" security seems like you could easily settle for sub-par opportunities, because you don't have a specific goal in mind and rather you have a relative criterion. There is no concrete goal.

 

(I work on optimization problems and this seems to be analogous to settling for local optima instead of looking for global optimum.)

 

My second criticism would be how this strategy works during an insane market bubble, or even an absurd market condition. Imagine a theoretical scenario when every opportunity is insanely overpriced or for some reason has zero or even negative yields! In those situations, it does not make sense to invest even on the "marginal" investments, because they would likely generate negative returns. I don't like the fact that you HAVE to be invested based on the pre-determined allocation percentages.

[Lupo Lupus]

• By conditioning on the valuation of my portfolio, rather than the market I can take into account that undervalued securities relative to the market are not equally available. For example, during the dotcom bubble the market overall was overvalued, while value still had a decent implied performance.

 

I didn't get this part - could you please clarify?

 

 

One approach (which I do not follow) would be to make the hurdle rate contingent on the general valuation of the market (ie, lower hurdle when market is richly valued to account for fact the opportunity set is less attractive). However, when the market is richly valued, the expected returns on value investing are not necessarily low at the same time (think of the dotcom bubble, where old economy stocks were priced OK). Thus my argument is that it makes more sense to condition on the expected return of your own investment style (aka value investing) rather than the general market when deciding how much to allocate to equity.

 

In general, I see a couple of limitations in this approach. The first big limitation is that it seems like this approach is less goal-oriented than the hurdle rate-based approach. In the hurdle rate-based approach, you set a specific and absolute target rate, which motivates you to find a very few best investment opportunities. I feel like this mind set makes you work really hard to find the best opportunities. The concept of "marginal" security seems like you could easily settle for sub-par opportunities, because you don't have a specific goal in mind and rather you have a relative criterion. There is no concrete goal.

 

(I work on optimization problems and this seems to be analogous to settling for local optima instead of looking for global optimum.)

 

 

I see your point on getting less ambitious because of the missing absolute hurdle. In terms of optimization problem, I would argue the other way around: by trying to find the most undervalued securities one considers the whole spectrum. If you use hurdle rates, you may buy a security with expected return of 11% (if your hurdle rate is, say 10%) while at the same time securities with higher expected returns are available. This should particularly a problem when prices are severely depressed and many securities pass the hurdle rate (eg, GFC!). The hurdle rate becomes then fairly meaningless.

 

My second criticism would be how this strategy works during an insane market bubble, or even an absurd market condition. Imagine a theoretical scenario when every opportunity is insanely overpriced or for some reason has zero or even negative yields! In those situations, it does not make sense to invest even on the "marginal" investments, because they would likely generate negative returns. I don't like the fact that you HAVE to be invested based on the pre-determined allocation percentages.

 

Good point -- thank you for your thoughts Clutch!

[clutch]

Regarding the hurdle rate becoming meaningless - you could resolve this issue by treating the hurdle rate both as a hard constraint (something you must satisfy) and considering it also as an objective (greater from the hurdle rate the better). And say, you also put a constraint on the maximum number of securities you'd hold at a time. During the opportune market, if you find a number of favorable investments above the hurdle rate, you would also need to rank them and select n best securities. (I guess at this point the strategy becomes more like yours).

 

Anyways, thanks for sharing your interesting idea.