Investing in an Index Portfolio

Over Thanksgiving, I had a few conversations with family about what and how they should be investing. Investing in an index portfolio is the easiest thing to recommend to people. In this post, I outline how I’ve thought about it and my current plan for investing in an index portfolio

Goal: Create a portfolio of indexes that we can invest a significant amount of capital, potentially the majority of your net worth, into and hold for the next 10+ years.

Don’t Reinvent the Wheel

Below are the portfolios that Wealthfront, Betterment and the Vanguard Target Fund invest in on behalf of their clients.

YTD Return and Yield are as of May, 6 2017


On average, the 3 portfolios have appreciated 6.96% this year, distributed 2.25% in dividends and carry an expense ratio of 0.12%.

Although the specific indexes are different, the allocation in broader categories is similar.

I like Betterment’s portfolio for 2 reasons.

1. Granularity. Unlike the Vanguard portfolio, that invests in a Total Stock Market index, I like the granularity of the Betterment portfolio. Although I don’t have a preference for large or mid-cap stocks at the moment, I like having the option of being more specific with our allocations if need be.

2. Lowest cost, highest return. Even though past performance isn’t an indicator of future performance, Betterment’s portfolio has the highest YTD return (7.12%), highest yield (2.41%) and the lowest expense ratio (0.10%).

Buying

Obviously you don’t want to buy your entire position all at once, especially since the market is on the expensive side. Instead we want to develop a plan to slowly buy while minimizing the average cost.

To get a better understanding of how to do this, let’s look at the historical prices of VTI (Vanguard Total Stock Market ETF) before and after the recession in 2008.

The first thing that jumps out at you is how much the market has appreciated since the recession. We’re way past the peak in 2008. We can debate whether this is sustainable but the stock market has appreciated roughly 7% over the past 100 or so years so if you zoom out far enough, it basically goes up and to the right.

Important Dates

  • High of $77.74 on 10/9/2007
  • Low of $33.70 on 3/9/2009
  • $136.07 on 12/02/2017

If you had bought your entire position at the absolute bottom for $33.70 on 3/9/2009 (my birthday!), today you would be up 303% in 8 years. Calling the absolute bottom is a fool’s game but it’s helpful to know what the max return is.

Correction and Rebound Velocity

What we’re interested in knowing is when a correction happens, and they do always happen eventually, how quickly does it drop. Similarly, how quickly does it rebound.

  • 0.25 years to drop 10% from High
  • 0.76 years to drop 20% from High
  • 0.99 years to drop 30% from High
  • 1.42 years to go from High to Low
  • 3.88 years to go from Low back to High

It took 1.42 years to go from the High to the Low and 3.8 years to climb back to the High. Depreciation is swift, appreciation is slow.

Furthermore, depreciation tends to decelerate as it approaches the bottom. The first 10% drop happened very quickly (0.25 years) but it took the rest of the year to drop 30%.

Backtesting

With general observations under our belt, let’s come up with some very simple plans and backtesting them to see how they would have performed during 2008. I think it’s a reasonable proxy because we should also be at the tail-end of this rally.

Let’s start by dividing the capital we want to invest into 20 parts. To be conservative, let’s say we start buying at the High of $77.74 on 10/9/2007. From there, I tried different simple rules for speeding up or slowing down the frequency of buys.

The plan #’s are the leftmost column.

Plan 1. If you had simply bought every 3 months starting at the absolute peak and held until 10/3/2014 (7.23 years), you would have returned 74.85%. Still pretty good.

Plan 2 to 5. If you were trying to time the market a little bit, you might buy every 6 months and accelerate your buying to every 3 months after a 10%, 20% or 30% correction. That would have bumped up your return to about 86.33% (20% correction).

Plan 6. We can do a little bit better if we have a tiered accelerated plan where we start buying every 6 months, every 3 months if it corrects 20% and every month if it corrections 30%. That plan would have returned 98.44%.

Plan 7. The same purchase plan as Plan 6 but I changed the sell date to May of this year, an additional 2 years. That significantly bumped up the return to 126.80%. It pays to hold.

Plan 8. The same purchase plan as 6 but I started it a year before the peak so we’re buying on the way up. Just like we can’t time the bottom, we can’t time the top. Turns out it’s not too bad returning 103.13%.

Recommendation

Similar to the plans above, I recommend we split our allocation into 20 parts and do the following:

  1. Buy 1 part every 6 months starting now
  2. If the market corrects 10%, buy 1 part every 3 months
  3. If the market corrects 20%, buy 1 part every 1 month

We want to hold this position for as long as 10 years, well into the next rally.

Further Work

I’m planning on doing more work on the backtesting results. This was an initial investigation. Specifically I’m interested in doing the following.

Backtesting the Betterment Portfolio.

The backtesting I did was on VTI, a Total Stock Market ETF, not on the indexes in the Betterment portfolio. I might do the same analysis on the Betterment portfolio in the future but the outcome shouldn’t be significantly different. Certainly on the risk side, the Betterment portfolio should be safer because of it’s greater diversification.

Backtesting through several corrections.

I only looked at the 2008 correction but to be thorough I could look through all the corrections in the last 30 years.

I also haven’t done extensive research into how other people are tackling this, I’d like to do that.