The recent ETFDB newsletter profiled “3 ETF Types For Downside Protection”. RBS’s family of ”TrendPilot” exchange-traded notes offer investors exposure to a dynamic trading strategy applied across a variety of asset classes. Each fund tracks a rules-based index that strategically shifts allocations based on a simple historical moving average. The strategy is relatively straightforward: when an ETF’s underlying index is at or above its 100-day simple moving average, the fund goes long the underlying securities. If, however, the index closes below the simple moving average for five consecutive sessions, the ETF will shift exposure to “safer” short-term U.S. Treasuries. China Trendpilot ETN (TCHI)US Large Cap Trendpilot ETN (TRND)US Mid Cap Trendpilot ETN (TRNM)Gold Trendpilot ETN (TBAR)NASDAQ-100 Trendpilot ETN (TNQD)Oil Trendpilot ETN (TWTI) http://etfdb.com/2012/etfs-for-downside-protection/#more-584...The idea is attractive. An investor would own the underlying when it’s doing well, and he/she be put into cash when it isn’t. So I got to wondering how such an idea might work out if it were applied to bond funds. But, first, as with any experiment whose results one wants to apply, the experiment itself has to be replicated. It’s safe to guess that their large-cap trend fund, TRND, would use the SP500 index as the underlying. So I plotted SPY (as a way to pick up dividends) against TRND and applied a SMA100. As you can see from the linked chart, the fund sat in cash from early Aug 2011 to late Dec 2011. OK, so far, so good. The fund seems to be doing what they say they would do. When a crossover occurs and persists, they get out. http://finance.yahoo.com/echarts?s=SPY#symbol=spy;range=1y;c...So the next question becomes, “Why a 100-day moving-average (and not other parameters?) and why a simple moving-average (as opposed to other types)? The answer to the latter question is easy. Simple moving-averages (SMA) are the easiest to calculate, and they also are very robust, besides being able to closely replicate the behavior of the other common types (exponential, front-weighted, triangular, etc.) simply by varying the parameters of the SMA until it closely matches the EMA, WMA, etc. So, on their choice of moving-average type, I have no quarrels. SMAs are plenty good enough. What I do have quarrels with is their choice of parameters. So I started running tests which you should duplicate. E.g., instead of 100, I plotted 90, 100, and 110 and saw no significant differences in the signals generated, as should be expected. But widening the parameter set to 50-100-150 did begin to make a difference, as should be expected. http://finance.yahoo.com/echarts?s=SPY#symbol=spy;range=1y;c...So the next question becomes, “How fast do you want to get in/out?” meaning how much delay in the signal set are you willing to tolerate and/or by how much are you willing to let prices get away from you, or move against you, before you act? Hold that thought in mind and take a look at another of their funds, the one that tracks the price of gold. http://finance.yahoo.com/echarts?s=GLD#symbol=gld;range=1y;c... In this case, a 100-day SMA does a very sucky job of creating acceptable entry/exit signals, but a 25-day SMA does a pretty efficient job of capturing the bulk of the available moves. Obviously, a lot more testing would need to be done with various indexes and parameter combinations. But two conclusions already emerge: (1) Their funds can be done better by individual investors and don’t need to be bought. (2) Every index is going to have its own personality and needs to be traded on its own terms. Charlie
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS. Market data provided by Interactive Data.
Company fundamental data provided by Morningstar. Earnings Estimates, Analyst Ra