We’ve Updated Our Asset Allocation Models

Here at Kruse Asset Management we are constantly striving to improve all facets of our client experience, including our communication, superior technology and services, and of course, how we manage client portfolios.

When it comes to investing, there are three basic areas to focus on: stock selection, market timing and asset allocation.

Most of the time people seem to focus on which stocks they should be buying and selling. Then when the market shows signs of weakness, their focus shifts to market timing; more specifically: should we be getting out? However, famous financial studies (if there are such things as “famous financial studies”) done by Gary Brinson, and then followed-up by Roger Ibbostson & Paul Kaplan, have demonstrated that asset allocation is more important than both of the other areas… by a factor of ten!

If you want to understand your portfolio’s performance, including risk and returns, the often overlooked area of asset allocation is where you should be spending the majority of your focus.

To that end, with new technology, here are some areas that we’ve been able to improve upon with respect to portfolio management for our clients:

1. Correlations and Covariance (Asset Allocation)

Without getting too granular, the basic principle of Modern Portfolio Theory is that every time we add an investment to a portfolio that does not move in lock-step with the rest of your investments (measured by covariance), that portfolio can potentially have higher returns and less risk! That is a win/win! The more varied in their movements that the investments are from one another, the better.

However, markets change. The relationships of investments (covariance) change over time, as can be seen in the chart below.

For example, rising interest rates could be a headwind to the economy (bad) or it could signal an improving economy (good). The same could be said for the price of oil: higher prices could be an added expense on people, as well as manufacturing and transportation companies (bad), or it could signal greater demand for those products and a more robust economy (good).

Conveniently, the “experts” on TV can explain stock market movements after the fact; unfortunately, they are actually worse at predicting the relationships than a coin flip would be!

With the goals of reducing risk and improving potential returns, each additional investment and asset class in your portfolio increases the number of calculations and complexity exponentially. Because these relationships all change over time, understanding the risk of your current portfolio in the present was extremely challenging... until now.

New technology has enabled Kruse Asset Management to build self-updating models that incorporate recent relationship changes into your personal asset allocation so that we can ensure that your goals are being met and your risk tolerance is being honored within the context of the most recent, up-to-the-minute investment information.

For example, our models can tell when an asset class, like commodities, is acting more like a stock than a bond (and to what degree). How are commodities related to the stock market (is oil helping or hurting your investments?) Are “correlations going to 1?” which is a phrase that was thrown around during the financial crisis of 2008 that implied diversification did not help during the down-turn. On a side-note: correlations never really got close to 1, a perfect correlation; and historically, high levels of correlations are more prevalent during up-markets than down.

These questions—and many more—all affect the risk of your portfolio. Are we over our “risk budget” through no fault of our own, other than some asset classes lost some diversifying benefits that we couldn’t have caught unless we re-ran all the correlation and covariance calculations in real time (hundreds of thousands of semi-manual calculations)?

Now WE know...and we don’t know of any other firm (big or small) that is doing this.

The big firms might have the expertise, but given the level of compliance and legal hurdles, potential clients’ portfolios need to be sent out each and every time to run their outdated, but approved, allocation models and simulations of old. Our CEO plays basketball with someone at a large investment bank and he quipped that if he gets out of bed to get a drink of water in his own house, compliance is waiting for him to check out his glass.

Small firms might have the flexibility, but rarely have the technical ability to build these types of models, and if they did, are generally so busy looking for new clients that they just don’t have time to build them. Additionally, the general feeling is that it really isn’t worth their effort because their clients don’t know the difference anyway. So they often farm out much of their investment decisions to off-the-shelf software, mutual fund managers, or worse, have no real plan other than their own gut-feelings (usually a really bad idea).

This exercise had the added benefit of revealing true nature of some asset classes. Investments that were supposed to be acting one way given the mandate of the strategy, but now we have the data that shows that they were acting just like stocks, bonds or even cash... all the time. Since some strategies provided no real diversification benefits, we were able to not only simplify our models, but make them better and more robust with fewer moving parts.

2. Tactical Shift Adjustments (Timing)

In light of the recent market pull-back and increased volatility, it seems particularly pertinent to discuss a good time (if any) to sell when the markets are weak. More often than not, the correct answer—even in the face of a double-digit decline—is to stand put, make no changes and do nothing. Why? Because getting out is the easy part... it is the re-entry point that is challenging, due to something called the “anchor effect.”

Say the market is showing real signs of trouble, which usually takes at least a 10% decline from a recent peak, and so we get out. But when do we get back in? Most don’t get back in while the market is still falling—we got out because it is falling, so it would be hard to get in because it is still falling. So the market has to show signs of recovery, right?

If the market rebounded 10%, that might be a good signal the market has turned around, but now it looks too expensive. Mentally, you’re “anchored” at the market lows, knowing that you had an opportunity to buy at much lower levels, so ironically you need wait for another pull-back. If that pull-back comes, then you might tend to second-guess the recovery. If it doesn’t come, you don’t get back in either because the markets are even more expensive now. Either way, by the time you’re convinced that the market has recovered and that you must get in now or lose more up-side, the round-trip of the exit and re-entry usually costs more than had you just stayed in the entire time.

That said, we did the research and found there is evidence that suggests if the markets pull-back 20% and enter “bear market” status, it might be a good time to sell. The re-entry point would be a 10% rebound off the lowest point. An average “bear market” is down 35%, so selling at 20% still allows some room to the downside, but also gives us a reasonable re-entry point such that we can miss some volatility, and even come out on top (as seen in the chart below).

Back-testing 50 years of data suggests getting out sooner than a 20% drop is likely to be counter-productive. As of the time of writing, our max drawdown is only 13% off our recent highs.

So when an asset class enters “bear market” territory, our models will shift at least 50% of that asset’s allocation into cash until an appropriate re-entry point is reached. This has the added benefit of reducing the portfolio’s overall volatility as well.

3. Relative Strength (Investment Selection)

There has been quite a lot of research that show that relative strength can be a good indicator of future performance in asset classes, market segments, countries’ performance, etc.

What though is Relative Strength?

The Super Bowl just took place, and if I were to ask you which group of teams you think would do better next season: the teams that made the playoffs this year or the teams that didn’t, you’d probably pick the teams that made the playoffs this year.

I might ask you why you think these teams will do better next year and I’m sure that you could come up with some theories (faster, stronger, more talented players, better coaching, more cohesiveness, etc.), and you’d be right, but you wouldn’t really know to what extent these all played a role because it is impossible to understand all of the interactions. However, in the context of the original question, you don’t need to know why—you only really need to know that those teams are better now and will probably be better (as a group) next year. That is Relative Strength.

Sure, some teams that made the playoffs this year won’t next year, and teams that missed will be in contention next year, but if we were given the chance to bet, straight-up, we’d keep betting on the play-off teams until they stopped making the play-offs.

Unfortunately, Vegas does not give us that straight-up bet, but the stock market does every day. And while we used relative strength in the past, the process was not able to be incorporated, real-time, into the models... until now. Now as asset classes, sectors and segments move in and out of levels of high and low relative strength, our models adjust in real time following the money flow. Just like in football, we don’t really have to have all the answers as to why, we just need to know it is happening now, and we’ll continue placing bets on teams that made the playoffs until they stop making the playoffs.

All of these changes should make for portfolios that are simpler, take on less risk, are more aligned with client risk tolerances, and yield better long-term returns. We’re excited to implement this new change and look forward to seeing performance improve even further heading into the near future.

If you would like to know more about our asset allocation model then please do get in touch—we’ll be happy to discuss it with you!