In this post, we explain what back-adjusted market quotes are, why we require such quotes to calculate our tactical portfolios, and which implications the use of these quotes has. This insight is not only meant to satisfy your curiosity but will also help using TuringTrader better.
High-Level Structure of a Tactical Portfolio
TuringTrader.com tracks more than forty tactical strategies. And even though these follow very different investment approaches, namely trend-following, momentum, mean-reversion, and volatility-targeting, they all share a very similar high-level structure:
- Consume market quotes and economic data
- Calculate indicators
- Determine investible assets
- Rank investible assets and determine portfolio holdings
- Calculate orders and position sizes
- Execute orders
We can see that data are at the core of every algorithm. And it is easy to understand that when we consume incorrect data, none of the steps that follow make sense. Or in other words: Garbage in, garbage out.
Further, pretty much all tactical strategies perform an asset ranking of some sorts. Such ranking is performed on one or more metrics, typically indictors we calculated. A key requirement to successful ranking is that the metrics we are looking at are comparable.
Corporate Actions, Dividends, and Interest
As an example, momentum strategies calculate and rank price momentum, with momentum calculated as the difference between the asset price today and at some point in the past. But an asset’s market price paints an incomplete picture. Let’s look at a few examples:
On June 3rd, 2022, Amazon completed a 20-for-1 split of their shares. To illustrate this further, AMZN closed on Friday 06/03 at $2,447.00 per share and opened on Monday 06/06 at $125.25. But this price change didn’t indicate a corporate meltdown. Quite to the contrary, the value of AMZN shares actually went up:
$2,447.00 / 20 = $122.35 < $125.25
Similarly, many stocks pay interest or dividends. On April 30, 2024, BIL closed at $91.79 per share and opened on Monday 05/01 at $91.43. Again, things are not what they seem. BIL paid interest of $0.39 per share, which makes up for the difference:
$91.79 < $91.43 + $0.39 = $91.82
Back-Adjusting of Quotes
The solution to this problem is to use back-adjusted quotes. With these quotes, the last price matches the current market price. However, historical prices are adjusted such, that the price history reflects the asset’s total return, including all splits, mergers, dividends, and interest payments. There is a fabulous article on the NASDAQ blog explaining the mechanics in much detail.
It is important to notice that back-adjusting rewrites the asset’s price full history. So, if a stock pays a dividend after market close today, today’s price and all prior prices will be adjusted.
Small Changes – Large Effects
Of course, these price adjustments shouldn’t make a difference. While the absolute prices changes, their relationship to each other stays the same. But unfortunately, reality is more complex than that. Because price data have limited numerical resolution, any adjustment made to the prices will result in miniscule changes in the indicators calculated from these prices.
Above, we identified asset ranking as one of the key elements of a tactical portfolio. If we imagine that under certain circumstances the metrics used for ranking are very close together, it becomes clear that such miniscule changes may result in our trading strategies favoring one asset over another. Such behavior is described by Chaos Theory and the example of the famous Butterfly Effect.
All charts and metrics published for TuringTrader portfolios are backtested. This means that all of the strategies decisions regarding assets and position sizes are based on the data available today. Due to back-adjusting of asset prices it is possible that metrics and asset allocations, both current and historical, may change. This does not happen often, but when it does it is disruptive: expected issues include changes to the equity curve, deviations in the strategy’s key metrics, unexpected fluctuations in the estimated value of accounts invested in the strategy, changes to the asset allocation, and unexpected rebalancing orders.
The Timing of Price Adjustments
In the context of asset price adjustments, it is also important to consider the timing. To ensure smooth trading operations, any such actions, stock splits, dividends, or interest payments, happen outside of regular trading hours.
This explains why TuringTrader cannot post new asset allocations immediately after the markets close. We have to wait until trading prices are finalized, any corporate actions or dividend/ interest payments completed, and our data vendor finished calculating and publishing adjusted data.
Conclusion
At TuringTrader.com we are committed to providing our members with a flawless service and a seamless investment experience. Nonetheless, we encourage investors to learn more about the mechanics of tactical asset allocation. Knowing how the sausage is made helps investors avoid frustration by anticipating issues and acting accordingly.