Highs and lows: a behavioural and technical analysis

Abstract

We find that turnover rises on n-day highs and lows and is an increasing function of n. We offer several explanations from the technical and behavioural finance literature for why traders might use these signals. Turnover is persistent following these events, and new lows provide abnormal returns for up to 6 trading days.

‘Technical analysis is about as useful as going to a fortuneteller, as far as I'm concerned. There is simply no evidence that these patterns mean anything…’ ¹ (Burton Malkiel, 2003).

¹ Interview in Portfolio, The National Association of Realtors, October/November 2003, available at: http://www.nareit.org/portfliomag/03sepoct/capital.shtml

Acknowledgements

We would like to thank the Withcomb Center for Financial Research for providing access to the CRSP data.

Notes

¹ Interview in Portfolio, The National Association of Realtors, October/November 2003, available at: http://www.nareit.org/portfliomag/03sepoct/capital.shtml

² I rely on www.stockcharts.com for these definitions, but these are widely utilized indicators in the practitioner literature.

³ Huddart et al. (2007) use ‘the the highest (lowest) daily closing stock price in the 52-week benchmark period ending 20 trading days … before the last day of the observation week’. We did not follow this construction.

⁴ The trading volumes on NYSE and NASDAQ market are calculated differently due to their different market microstrutures. A trade is counted only once when the NYSE specialist matches a buyer and a seller. A trade is generally double-counted on NASDAQ during our sample. If a buyer sells a trade to a market maker, and he then resells the trade to a seller, NASDAQ will record this as two trades.

⁵ http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html