Asset Class Breaks in Performance -

Strategic vs Tactical Investing

An Asset Class break exists where there is a gap in the performance stream. Meaning, the asset class has been divested to a value of $0, only to be re-invested at some future point in time, thereby creating an artificial 0% return stream, i.e a gap. Some firms have debated the issue of whether or not to a) link the broken return streams together with 0% returns as ‘filler’ or b) keep the return streams distinct as 2 different return streams, technically stopping and starting the return streams into 2 different, isolated periods. In order to grasp this complex topic, some color on a few topics should be discussed first.

• Strategic vs Tactical Investing

• GIPS rules on composite breaks in performance

• Intent of the Asset Class Groupings

• Geometric vs Arithmetic Linking

• Signed Investment Policy Statements and adherence to Stated Objectives

With respect to Strategic investing, one needs to determine specifically what return outcome is desired (these are considered the Investment Objectives), and how your firm will measure those results against a specified benchmark. On the outset, this is done through an Investment Management agreement or Investment Policy Statement. This document outlines the goal(s) of the accounts(s) over time, with defined objectives that list out the Tactics your firm will use to achieve the desired investment result/objective.

In Tactical investing, one is more or less ‘managing’ the account to a very broad set of investment goal(s). Tactical investing is basically the day-to-day and/or ‘fly by the seat of your pants’ type of investing. Therefore, the Strategy is the overall, but the Tactical is the methods by which your firm intends to implement the Strategy.

The main question is always, “How can the Strategic goals be accomplished within the designated limits and resources of authority?” The only way that can happen is to ensure that the Tactics utilized create results which lead to the Strategic benefits your firm desires. If there is no specific Strategy intended, then Tactical investing comes into play.

If an account is being managed to an overall objective, a reasonable benchmark should be utilized. If an account is being managed on a day-to-day basis, without regards to an overall plan, then a reasonable comparison is not available, i.e. no benchmark.

In order to determine how to handle Asset Class breaks, your firm should clearly define which group (Strategic or Tactical) its investment style is grouped.

If a GIPS compliant firm has a GIPS composite whereby all of the accounts have been removed, the firm has 2 options: a) consider this a closed strategy/composite and then terminate as such or b) consider this an open strategy for potential future investing. If your GIPS firm chooses B, then there are 3 rules that must be followed:

1. Returns both before and after the break must be presented

2. Returns for the partial year must be presented

3. Returns must not be linked across the break

If your firm thinks about Asset Class breaks in terms of GIPS rules for composite breaks, and applies this to its specific situation, consider these topics:

1. If an account has an Asset Class that has a break in performance but the Asset Class represents a Strategy that is continuing and intended to be implemented based off of an overall, pre-defined investment objective, then it could be assumed that your firm should follow the best practices of GIPS rules and follow the 3 rules mentioned above.

2. If an account has an Asset Class break that has a break in performance, but that Asset Class no longer represents an investment strategy that is desired by the client, then that strategy (ie Asset Class performance) would die and be considered closed, with the old track record being kept for historical purposes, but not linked to or represented to any future investment decision.

Please note that most investment management firms that manage accounts do not show performance at the Asset Class or Sector level, but just in Total. However, firms that do show the performance at such further defined levels, they do so because they are managing the assets such at those levels utilizing some sort of Strategic Asset Allocation (ie a model).

If your firm does not use models or manage accounts according to a pre-defined asset allocation, it would beg the question of why your firm would show asset class performance to begin with? It might not be useful to show Asset Class performance if the firm is not sticking to a defined Asset Allocation.

For Example:

Let’s say your firm has a manager that has a model called Equity Aggressive and this model has the asset allocation of 50% US Large Cap, 40% US Small Cap and 10% International. Therefore, this would be represented in a pie chart, and the stocks would be held according to the weights assigned in the Investment Policy Statement. Each Asset Class is assigned a benchmark, and then the Total account benchmark is a blend of those assigned weights.

If the manager is managing specifically to that pre-defined Asset Class, then there would be no breaks in performance for those Asset Classes, because the manager is managing to the intended strategy. If your firm does not manage to a model, then your firm would realistically not want to show Asset Class returns, because there is no pre-defined strategy. So, if a manager wanted to see his/her performance broken down into Asset Classes, then he/she should be managing to a specifically designated strategy.

Annualized returns help to analyze the differences in returns between any two investments, thereby holding the time period constant. Basically, annualizing ‘normalizes’ the effect of an investment by ‘smoothing out’ the volatility that may be shown in a performance report. Cumulative returns show the compound performance of an investment over long-term periods. Cumulative returns incorporate the assumption that investment gains are reinvested into the portfolio and compounded over time. Cumulative returns are worthwhile to show, however, they do not allow for normalizing over time.

In doing a calculation on the annualizing effect of a break in performance (click link to see example), one can see that in the General Sample Tab, the return through 3/31/2015 shows what could be represented as a continual asset class return stream resulting in a 79.59% Since Inception return. Then in April 2015, the asset class broke resulting in 0% returns for the next few months. Then in September 2015, the asset class was purchased back into, but, because of the 0% returns that are put into the historical stream, this has affected the annualization of the returns by ‘dragging down’ the previous historical performance. Therefore, when the performance picks back up again, it dragged down to 56%.21 (see that the cumulative was unaffected by the 0% returns). However in the 2nd example (Alt Example tab), if that performance re-starts, the Since Inception would start over, over the long-term, benefits the client overall with regards to the performance returns being shown.

Continuing from the Intent on Asset Class Groupings topic, if there was an adherence to a specific set of signed objectives set forth in the Investment Policy Statement, then there should be no asset class breaks. If the client agreed to a 60% asset allocation of international equity, then the client would always have around 60% in that asset class. If for some reason the client got out of that asset class, then there should be a newly signed IPS that would show the change in the client’s asset allocation. Therefore, if the allocation weight actually went to 0%, then there would be no need to link a potential asset class break together, because this allocation change represented a Strategic Asset Allocation move at the total account level out of that asset class.

The only way that linking of the performance would be an issue is that if the manager made a Tactical Decision to deviate from the IPS for a short period of time. Meaning, the manager decided that ‘in the short-term’, this asset class is going to tank, and instead of re-doing your whole IPS, we are just going to get out of it for the time being.” In this situation, you would link the returns, simply because the IPS stated you ‘should’ be in that asset class, but the manager made a Tactical decision to move out of it, temporarily. The benchmark would continue on, unbroken, because you would continue to evaluate the manager against the investment decision that was chosen in the IPS.

After reading through the sections above and meshing those ideas together, one should really look at ‘what is the intended strategy’ or ‘what is the intent of the investment objective’ of the overall account & what actually occurred in the account & why. Therefore, a decision tree could be implemented with the following options:

Option A: Show the performance with all of the old and new returns, separately, as 2 distinct return streams.

Example: Fund sold out in May 2015. Therefore, only as of April 2015 would the performance be shown for that asset class and only an April 2015 report would ever show that information as a solid, unbroken return stream. If then in December 2015 the asset class was purchased back into the account, the old information would appear back on the report through April 2015, with an additional line shown the new return stream starting December, 2015. Only would the historical return stream appear back on a report if NEW returns were generated for that asset class. Benchmark returns would also follow suit.

Option B: Leave the old performance as a separate, static return period, with no linking together.

Example: Fund sold out in May 2015, therefore, only information as of April 2015 would exist for that asset class. Then, starting in May 2015 that asset class information would not appear and would only ever exist if a historical report was run for the April 2015 time period. Once the break occurs, that old performance stream would ‘die’ and never get linked to the new information.

Benchmark outcomes in this scenario depend on what the investment objective was on the account. Benchmark outcomes could be:

1. Account is managed towards a pre-defined Strategy/Asset Allocation, manager deviates from the strategy, benchmark should continue so the manager can be evaluated against the investment decision. If manager buys back into the strategy, the performance returns are linked against an unbroken benchmark return.

2. Account is managed towards a pre-defined Strategy, and the manager is allowed to buy/sell in/out of asset classes (because this is the accepted Strategy), then the returns can be linked. The benchmark return starts/stops with the investment in the asset class (this is basically security level performance).

3. Account is managed towards a pre-defined Strategy/Asset Allocation, manager changes the strategy according to a specified mandate change, benchmark return stops so the manager can be evaluated against that investment decision. If the manager buys back into the strategy, the performance returns are not linked, and performance is measured against a broken benchmark return, along with the broken track record.

Option C: Link the old and existing return streams with 0% returns in the ‘gap’ period, regardless.

Example: If a manager sold out of the fund in May 2015, the manager would only have trailing information through April 2015, and that return information would remain static and never change. That is until the manager re-purchased into that asset class. Once re-purchased into that asset class, the performance would then be updated and tacked on starting with that month’s performance, thereby changing the trailing return information, once the new information is populated.

With regards to this option, it is unclear as to why a manager would have to link 0% returns together, if it just eventually drags down the annualized performance? It is assumed that showing a cumulative return in these instances would be warranted, but cumulative returns can be misleading.

With regards to consideration of the benchmark, if the decision was a Strategic one, then the benchmark would also go to zero. If the manager sold out of the asset class, then technically the manager would also need to make the benchmark zero. One must think about whether or not the decision was an Asset Allocation shift or Tactical decision. Conversely, the reason the benchmark would continue not go to zero, would be because the manager may have deviated away from the IPS on the account. If the decision was to sell out and sway away from the model, then the benchmark return should continue on, and the manager should be evaluated against that investment decision. If the client decided to sell out of an asset class, but according to the model the client was supposed to be in that asset class, then the manager would keep the benchmark intact, and just want to link the old and new return streams with 0% returns. If the decision was made without regards to the asset allocation model, then this would show the impact of that specific decision.