Fun with Fintech (part 2)

This is part two of a two part series. The previous installment can be found here, where I cover some background info on buy and hold strategies and other stuff.

Recap (copied from previous post)

Despite everything I wrote in my last post, I’ve never been 100% comfortable with buy and hold for my own strategy. I watched the 2001 and 2008 crashes in excruciating detail and saw many people simply fold when the market swung lower, dropping 20%, 30% and even further. Dollar cost averaging works extremely well in principle but it takes psychological resolve to buy into a market that is precipitously falling. Many simply can’t do it and so suffer the downswings without much of the upside.

So while I’m a strong believer in the S&P500 as the best approach to picking stocks, I’ve always wanted an investment product that would signal whether we’re in a market uptrend or downtrend and allow me to take appropriate action. One of the foundational beliefs of the stock market (called Dow Theory) is that it moves in trends. That is, when it goes up, it goes up for a while and similarly the reverse is also true. A trend could be defined as a series of higher highs, and lower lows. So while it’s not possible to predict or perfectly time the market, it should be possible to tell which way we’re trending and then to take informed actions based on this. From a personal perspective, my ideal would be to hold / buy the S&P500 while in an uptrend, and be out of the market altogether during long downturns.


Solving for a market of one

I’ve had a long held belief that the best way to create something of value is to solve for your own problem first. By taking this approach, you’ll have greater empathy for the (eventual) customer, have passion for the solution and have greater insight into the nuances that can make the difference between “good” and “great”. Of course, the potential downside with this technique is that you may find that you are the only customer. But even in this worst case scenario, you can find satisfaction knowing that you improved your own quality of life and in doing so, brought value into the world.

This was my initial intention with this research — find something that works for me. What I wanted was a system that :

  • Didn’t require me to have to constantly watch the market. I wanted something that I could mostly fire and forget and only needed me to check back once a month or so
  • Was optimized for a long term strategy
  • Didn’t incur significant fees or incentivize bad behavior (I’d define bad behavior as overreacting to good or bad news or trading too regularly)
  • Didn’t require me to pick stocks. We already established in my previous post that the markets are the best stock pickers, so I wanted to use the S&P500 and similar indexes for this
  • Reduced losses during bear markets. For this I was willing to reduce the % gain in good times, in order to reduce the % losses in bad times

Once I established my own needs, I began to chat to others and found out this resonated. After a number of conversations, I was able to generalize the following higher level, human needs:

Lets build a system that solves for this

Momentum

My research led me to the study of “momentum” which can be broadly defined as the degree to which a stock (or market) is moving in any particular direction. There are different ways to calculate this, but a simple one is to look at the gain in stock price over time. Other factors like trading volume can also be considered, and there’s been additional derivative signals created around this. This area looked promising and after a while I came upon research by Gary Antoacci who details a momentum based approach in his book Dual Momentum Investing. This served as the early inspiration for my own approach.


Introducing Rivolv

Building on the momentum research, I built an algorithm with the intention of satisfying all my needs. I call this “Rivolv” and this is how it works.

At the beginning of every month, Rivolv will make one of two possible recommendations:

  1. BUY. If a buy signal is generated, it will be accompanied with a recommended index to buy. In this case, I move all my investable position into the recommended index
  2. SELL. Conversely, in the case of a sell signal, I’ll sell all my positions and move entirely into cash.

To illustrate this, image a basket of ETF’s — we’ll take SPY (the S&P500 index) and IWM (Russell 200 index). The following illustrates what happens over the course of several months:


What are the benefits

It seems dramatic at first — the idea of moving entirely into a position, but the reality is that these signals are triggered quite infrequently — in some cases a sell signal is not generated for over an entire year. In practice, it feels intuitive — there’s real appeal with being in the market when it’s going well and being out when it’s going bad. There’s a few benefits with this approach:

  1. Provides YOY returns close to the best indexes during bull markets. Because you’re not stock picking, you get benefits that are close to the best indexes (which by association is better than most funds)
  2. Protects against the substantial losses that a buy-and-hold strategies experience in downturns by moving into cash during bear markets
  3. Removes the fear of missing out by reentering the market during recoveries

How has it performed?

So this is the real question. While I built this model a while back, most the time since was spent monitoring and testing it. A few months back, I’ve started using it for myself. Fortunately, there’s a wide availability of stock market and dividend data, so it’s easy to test against historical data — a quant technique called backtesting. Here’s how this model performs against the S&P500 index over the past 10 years:

Note these are total returns, assuming reinvestment of dividends. You can kinda see how it’s working in this chart. During the crash of 2008, the model triggered a SELL signal which would have resulted in a move into cash. Then when the market picked back up again, it triggered a BUY and a reentry into SPY (the S&P500). Because there were smaller losses during the downtown, there was more to reinvest resulting in an overall higher performance. This is how Rivolv out performed over a 10 year period.

Here’s what the total and annual breakdown looks like:


Are there any drawbacks?

Yes, a few which are worth considering:

  • Back testing is only so effective. Past performance can’t be a guarantee of future gains
  • It’s conceivable that the behavioral psychology that underpins momentum in the market may simply stop working some day
  • In shorter time periods, this approach will always lag behind the S&P500 (kinda by design)
  • Moving “all in” on a position can feel psychologically difficult for people
  • There’s tax advantages to a permanent buy-hold strategy, something beyond the scope of this article

What’s does the current signal say?

Well, we just passed the start of the month (I’m writing this on 1st May 2017). Currently Rivolv is showing a BUY for SPY which is good for the month of May.


What are your plans with this?

Now that my first goal of solving for my own problem is completed, I‘m quite content with where things are at.

Moving forwards, if there’s interest from others, I’ll publish a monthly update on my blog with the latest signal recommendation. I also have a site I built to research product / market fit a few months back. This has more information on back testing and historical results. I’ll be happy to share this with you if you email me.

As for me, I’ve now locked plans for my next big venture and it won’t be in Fintech. However, I continue to find this space interesting so I’ll follow it closely and may fund more work with Rivolv in the future, or even just open source the approach.

Feel free to get in touch with any questions, requests, etc. For now that’ll wrap up this series — hope it was a fun read and informative. Oh, and more info on my next steps coming soon.


Disclaimer

This post is for education and informational purposes only and is not general or personal investment advice. Under no circumstances does this information represent, nor should it be interpreted as, a recommendation to buy, sell or hold any security, index, ETF or fund. Be aware that all investments carry risk. Investments may lose significant value. Users of these sites should consult an independent investment advisor before conducting any investment. Anyone using this site to inform their own investment choices does at their own risk and assumes all responsibilities for their choices and actions.