Elias Escreveu:Acho sempre alguma piada a ver a forma como os pequenos investidores se amontoam para vender nos números redondos.
Não são apenas os pequenos investidores.
Technical analysts associate these signals with two main predictions:
(1) Down-trends (up-trends) tend to be reversed at predictable support (resistance) levels;
(2) Trends gain momentum once these predictable support and resistance levels are crossed.
...execution rates tend to cluster, with particularly strong clusters at round numbers. Most critically, the pattern of clustering differs in important ways between take-profit and stop-loss orders.
The difference in clustering patterns between stop-loss and take-profit orders, which is wellknown
to market participants, is potentially important for our understanding of exchange-rate
dynamics.
Take-profit orders should tend to
reflect or reverse existing price trends, while stop-loss orders should tend to propagate or intensify
them.
To explain the first prediction of technical analysts regarding support and resistance levels –
that rates tend to reverse at these levels – I note that take-profit orders are more strongly clustered at
round numbers than stop-loss orders. This, combined with the reflecting effect of take-profit orders,
suggests that round numbers should act as partially reflecting barriers.
To explain the second prediction of technical analysts regarding support and resistance levels –
that rates trend strongly after crossing such levels – I note that stop-loss orders have a pronounced
tendency to be placed at rates just beyond the round numbers.
Since comparable orders data is generally unavailable to exchange-rate researchers, it is not
possible to prove that these patterns are truly representative of all conditional orders received by all FX dealers. However, there is no reason to suppose that these patterns are not representative. The bank in
question deals with the full spectrum of customers—bank and non-bank financial institutions, central
banks, non-financial corporate customers, wealthy individuals, hedge funds, commodity trading
advisors (CTAs)—from all over the world. Further, similar clustering patterns have been found in
large samples of limit orders for stocks: when stocks were typically priced in eighths (a practice that
ended completely in April, 2001 with the advent of decimal pricing), there was a strong tendency for
limit order prices to cluster at even eighths, and a stronger tendency to cluster at whole numbers
http://www.newyorkfed.org/research/staf ... /sr125.pdf