It can be important to delineate the results generated from the use of a 200 period moving average ( daily prices ) vs. a 10 period moving average ( monthly prices ). There can be more transactions and whipsaws in the daily series over a large test sample. The use of the monthly "contains" some of these whipsaws ( thus they would not be actionable signals ). Faber doesn't cover this topic in his paper and this was a topic of discussion in the late 80's and 90's when 30 and 40 period moving averages ( weekly prices ) were described by the likes of Stan Weinstein and others vs. the Fabian Switch Hotline and Investor's Daily as both implemented the 200 period MA ( daily ) in their services ( Fabian using it as a primary signaling strategy and IB using as a data visual ).
Additionally, it is difficult to use moving average cross heuristic ( or any other empirically derived tool ) and applied exclusively to an equity series as an isolated signaling tool towards generating meaningful alpha. At the very least, we have found that using the 10 period ( monthly prices ) moving average heuristic applied to both the equity series and the bond series increased returns / produced excess, risk mitigated alpha. An additional concept of investment in specific equity style indexes ( such as Nasdaq 100 index ) garnered further returns. http://tinyurl.com/jgg6ble

The use of the strategy applied to the Japan experience 1988 - 2014 http://tinyurl.com/ors89u3 and Europe 2006 - 2016 http://tinyurl.com/zdmw4ht

The use of CAPE valuation metrics in combination with the 10 period MA strategy/heuristic described has been an added feature in explaining alpha and risk mitigation during "high valuation" periods . http://tinyurl.com/pyq7my7