Rethinking Risk Management: A New Approach To Portfolio Construction

by: David Pinsen

Rethinking Risk Management

At its most basic, investing comprises two activities: buying securities we expect will generate decent returns, and limiting our risk in the event we are wrong.

Rule number 1: never lose money. Rule number 2: don't forget rule number 1.

Warren Buffett's advice about not losing money is well known, but harder to implement. It's impossible to invest without risking any losses, but typical investment portfolios -- ones constructed using a handful of broad-based index funds, such as the SPDR S&P 500 Trust ETF (NYSEARCA:SPY) often expose investors to the risk of large losses while offering low potential returns. SPY, for example, suffered about a 50% drop from October 2008 to late February 2009, and long term investors' reward for accepting the risk of such a large loss has been a ten year average annual return of about 3.5%.

I prefer to keep all my eggs in one basket and watch it carefully

Typical investment portfolios don't follow this bit of Buffett advice well either. They put their eggs in multiple baskets (such as SPY), ostensibly to limit risk. There are two problems with this sort of over-diversification. The first is that it doesn't protect against systemic, or market risk. When the market crashes, as it did in 2008, nearly all stocks plummet, and most other asset classes decline as well, as gold, for example - and ETFs that track it, such as the SPDR Gold Trust ETF (NYSEARCA:GLD) -- did.

The second problem is that high diversification dilutes potential returns. Instead of having your money concentrated in a handful of investments with the highest expected returns, you have it diluted among many investments with sub par prospects.

A New Approach To Portfolio Construction

If broad diversification doesn't protect against systemic risk, and dilutes potential returns, perhaps it's worth considering a different approach. I'll summarize one such approach here, and then drill down on the steps involved and offer a couple of sample portfolios. In summary, here is what I call the hedged returns approach:

  1. Find securities with high expected returns.
  2. Find securities that are relatively inexpensive to hedge.
  3. Buy a handful of securities that score well on the first two criteria; in other words, buy a handful of securities with high expected returns net of their hedging costs (or, ones with high net expected returns).
  4. Hedge them.
  5. Sell the underlying securities shortly before their hedges expire.
  6. Repeat.

The potential benefits of this approach are twofold:

  • if you are successful at the first step (finding securities with high expected returns), and you hold a concentrated portfolio of them, your portfolio should generate decent returns over time.
  • If you are hedged, and your return estimates are completely wrong, on occasion -- or the market moves against you -- your downside will be strictly limited.

Implementing This Approach

We've developed a portfolio construction tool at Portfolio Armor to automate this approach, and we'll show a couple of samples below, but first, let's look at how you can implement this approach on your own.

Manual Implementation

For the first step above, finding securities with high expected returns, one resource is the site you're reading now, Seeking Alpha. For example, in a recent article ("Regis: A Turnaround Play With Asymmetric Upside of 50+%") contributor Alpha Gen Capital wrote about a stock idea with a 50% potential upside. Now, if you agree with Alpha Gen Capital's analysis and expect Regis (NYSE:RGS) to appreciate 50%, that doesn't necessarily mean it has a 50% expected return for our purposes. Our approach includes hedging, so, when calculating expected returns, we need to consider the time period of our hedge. If you are going to buy protection on RGS for six months, then you need to consider the cost of that protection relative to how you estimate RGS will perform over six months. For the purposes of this example, let's assume you believe RGS will achieve half of its potential upside over the next six months. So, your expected return for it would be 25%.

The next step would be to find a hedge for RGS and see how much it costs. You can use your own method to do this, or you can use the Portfolio Armor hedging app, but either way, you'll want to decide what the maximum downside you're willing to risk is. For the purposes of this example, I'm going to assume you are willing to risk a 15% decline in RGS, but want to limit your risk to no more than that, in the worst case scenario. To use our shorthand terminology, 15% would be your "threshold". As of Wednesday's close, these were the optimal put options to hedge 1000 shares of RGS over the next several months against a greater-than-15% decline:

Note that the cost of this hedge, as a percentage of position value, was 10.28% -- quite expensive, but in this example, you also expect RGS to appreciate 25% over the next six months. This hedge actually expires a little bit more than 7 months out, so you may expect RGS to appreciate more than 25% over the time of the hedge, which means that your expected return net of hedging costs would be perhaps greater than 15%.

Now, you may find 15% to be too large of a threshold for you to risk, and you may consider paying more than 10% to hedge RGS far too much. The two examples below of hedged portfolios created with our automated tool will address both of those concerns, but the RGS example above gives you an idea of how of how you can implement this approach on your own. The key point here is to consider what your expected returns will be net of hedging costs. You may determine that another security has a lower expected return than RGS but a higher net expected return, because it has much lower hedging costs. If two securities have the same net expected returns, you'll probably want to pick the one that's cheaper to hedge. That's what our automated portfolio construction tool does.

Automated Implementation

Consider, for our first hedged portfolio example, an investor with $500,000 in cash who wants to maximize his potential return over the next six months while limiting his downside risk to a maximum drawdown of no more than 8%. For the purposes of this example, we'll assume that this investor doesn't have a list of favored securities to start with, so he'll leave the first field below ("Tickers") blank, and enter "500000" in the "Dollar Amount Of Portfolio" field and "8" in the Threshold field. If our hypothetical investor did have ideas he found from Seeking Alpha or elsewhere, he could enter their ticker symbols in the first field below, and he would then be given the choice of entering his own expected returns for them, or letting Portfolio Armor use its expected return calculations. Every trading day, Portfolio Armor calculates expected returns for more than 3,000+ stocks and ETFs. These estimates are based on analysis of historical returns as well as option market sentiment, which provides a forward-looking element.

Note that the descriptions of the fields above include question marks - if a subscriber hovers his mouse pointer over them, he'll see explanatory text, such as the one here explaining the strategy field.

After the few minutes it took Portfolio Armor to sort through and analyze the necessary financial data, our investor would have seen this result had he created this portfolio at the close on February 5th, 2014 (results could, of course, differ at different times based on market conditions):

The underlying securities in this portfolio are hedged in one of three different ways. The first, Biogen (NASDAQ:BIIB), is hedged as what we call a cash substitute. A cash substitute is a security that, when hedged with an optimal collar with a cap set at 1% or the current 7-day yield on a leading money market fund, whichever is higher, has a maximum downside risk less than or equal to the investor's threshold, a low hedging cost, and a net expected return greater than the current 7-day yield on a leading money market fund. The idea is to get better-than-cash returns while limiting portfolio risk per the investor's specification. The cap is the level of appreciation beyond which the underlying security will be called away: in a collar, you are selling someone else the right to buy your underlying security if it appreciates beyond that cap; the income you get from selling that right (the call options) offsets the cost of your downside protection (the put options).

The next three securities in the portfolio, CF Industries (NYSE:CF), Consol Energy (NYSE:CNX), and FedEx (NYSE:FDX), are hedged with optimal collars with the cap percentages set at their six month expected returns (the reason the numbers in the expected return column are slightly lower for CNX and FDX is that these have been adjusted downward because their hedges expire in less than six months; the assumption here is that investors will hold these underlying securities for six months, until they are called away, or until shortly before their hedges expire, whichever comes first).

The next three underlying securities, Lockheed Martin (NYSE:LMT), Northrop Grumman (NYSE:NOC), and Raytheon (NYSE:RTN) have no caps, because they are hedged with optimal puts instead of optimal collars. If you select the "maximize return" strategy, Portfolio Armor will consider both securities hedged with puts as well as securities hedged with optimal collars, ranking them by their net expected returns. In both cases, it will seek to include the securities with the highest net expected returns given your risk tolerance and portfolio size (since it presents only round lots, to minimize hedging costs, the maximum share price of securities the tool considers is limited by the dollar amount of your portfolio. For example, although (NASDAQ:PCLN) currently has one of the highest net expected returns of any security in Portfolio Armor's universe, the portfolio construction tool will not include it in a $100,000 portfolio, because $100,000 is not enough to buy a round lot (100 shares) of PCLN at its current price (more than $1100 per share).

Turning to the portfolio level data, we see that the maximum drawdown for this portfolio is 7.28%. That's the most this portfolio will decline in the worst case scenario. If each of the underlying securities above went to zero before their hedges expired, the portfolio would only decline 7.28%.

The potential return of this portfolio was 12.57%. That potential return is what the portfolio would return, net of hedging costs, if each of its underlying securities achieved its expected return within six months (or before its hedge expired, whichever came first).

Note also that the hedging cost of this portfolio was 4.6%. Let's say our hypothetical investor considered this hedging cost to be too high. For our second automated example, we'll have him enter the same dollar value of his portfolio, and same threshold, but select "Minimize Hedging Cost" in the strategy field. A few minutes after clicking the "Create" button, he would have been presented with this portfolio as of the close on February 5th:

A few differences you may notice in the construction of this second hedged portfolio. One is that there is more than one cash substitute included. The number of cash substitutes the tool includes varies as it fine-tunes its portfolio construction based on your inputted parameters and current market conditions. In addition to Biogen, which also appeared in the first portfolio, Canadian Solar (NASDAQ:CSIQ), Netflix (NASDAQ:NFLX), and Yahoo (NASDAQ:YHOO) each are hedged as cash substitutes. Weatherford (NYSE:WFT) is hedged with an optimal collar with its cap set at Weatherford's six month expected return. D.R. Horton (NYSE:DHI) is included twice, once hedged as a cash substitute, and once hedged with the cap set at the stock's six month expected return. Interestingly, the cash substitute DHI position has a higher net expected return despite having a lower cap -- the difference here is that an investor is effectively getting paid more to hedge DHI as a cash substitute. The cost of hedging DHI as a cash substitute is -4.23%, versus -0.71% when you hedge DHI the other way.

Turning to the portfolio level data, the maximum drawdown here is slightly higher than in the previous portfolio, but still below our investor's threshold. So, in a worst case scenario, if each of the underlying securities in his portfolio went to zero before their hedges expired, the investor would only be down 7.98%. His potential return with this portfolio, at 5.34%, is lower than the potential return of the first portfolio, but that's to be expected given that the strategy the tool used in this case was primarily to minimize hedging costs and only secondarily to maximize return: the hedging costs are also significantly lower in this portfolio -3.58%, meaning that the investor would effectively be getting paid to hedge this portfolio.

An Approach Worth Considering

An objection often raised against hedging is that it can be expensive. That's true, but the hedged portfolio approach is worth considering for a few of reasons we've touched on above:

  • The alternative, unhedged investing, offers uncapped risks and often modest long term returns (as in the case of SPY).
  • If you select securities with high net expected returns (high returns net of hedging costs), your potential returns can be attractive.
  • If you use a strategy to minimize hedging costs, you can reduce your cost of hedging and, in some cases (such as in the second hedged portfolio example above), effectively get paid to hedge.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.