{"product_id":"v174letters","title":"V.17:4: Letters","description":"LETTERS TO S\u0026amp;C\u003cbr\u003e\u003cbr\u003eThe editors of S\u0026amp;C invite readers to submit their opinions and information on subjects\nrelating to technical analysis and this magazine. This column is our means of communication\nwith our readers. Is there something you would like to know more (or less) about? Have you\nrun across trading techniques, services or products that have proved useful? Tell us about\nit. Without a source of new ideas and subjects coming from our readers, this magazine would\nnot exist.\u003cbr\u003e\u003cbr\u003eAddress your correspondence to: Editor, STOCKS \u0026amp; COMMODITIES, 4757 California Ave.\nSW, Seattle, WA 98116-4499, or E-mail to editor@traders.com. Letters published may be\nedited for length or clarity. The opinions expressed in this column do not necessarily\nrepresent those of the magazine.Editor\u003cbr\u003e\u003cbr\u003eMOVING AVERAGES AND CYCLES\u003cbr\u003e\u003cbr\u003eEditor,\u003cbr\u003e\u003cbr\u003eBrian J. Millards February 1999 S\u0026amp;C\narticle Moving Averages, First Principles\non moving averages fails to address\nimportant issues that arise when\nmoving averages (MA) are used in cyclical\nmodels. E. Slutskys 1937 article\nEconometrica told us that the moving\naverage transformation may induce the\nappearance of cycles where none exist\nin random data.\u003cbr\u003e\u003cbr\u003eOne practical approach may be to use\na second, different transformation to\ndetermine if a cycle with the same characteristics\n(wavelength, amplitude,\nphase) can be discerned in the data.\nExamining the closing price at the end\nof each week or the average closing\nprice (or first differences, as Millard\nsuggests) in each week are possible\nalternatives. Another strategy for those\nwith deeper mathematical skills is to\napply power spectrum analysis, which\ndecomposes the variation in data into\nthe sum of cycles, each with differing\ncharacteristics. Spectral analysis may\nconfirm the existence of cycles with\ncharacteristics similar to those shown\nusing the MA transformation.\u003cbr\u003e\u003cbr\u003eFurther, cycles sometimes appear in\ndata transformed by MAs because of\njust a few outliers or extreme values\nabove and below the trendline. Plotting\nthe raw data against the moving\naverage data may reveal whether just a\nfew datapoints are involved or whether\nmore pronounced, smoother cycles\nexist in the raw data. If in fact the\ncycles are caused by just a few outliers,\nthen entry and exit strategies must be\nmore carefully considered. To the extent\nthat more pronounced cycles can be discerned\nin the untransformed data, timing\nmay be easier.\u003cbr\u003e\u003cbr\u003eROBERT PHILIP WEBER, Ph.D.\nvia E-mail","brand":"store-traders-com","offers":[{"title":"Default Title","offer_id":52133979652465,"sku":"\\V17\\C03\\04LET.PDF","price":2.95,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0900\/3843\/6209\/files\/unnamed_0eeb41d4-6fd7-43a6-8284-bf8c4b9df235.jpg?v=1739660167","url":"https:\/\/shop.traders.com\/products\/v174letters","provider":"Stocks \u0026 Commodities","version":"1.0","type":"link"}