Go to Jeff Salisbury LWGate Home Page.From: "Katherine Malm" <kmalm@earthlink.net>
Subject: Re: [CANSLIM] CANSLIM Hunting List 09/20/02
Date: Wed, 25 Sep 2002 13:50:04 -0500
Next Article (by Author): [CANSLIM] Low Number of New 52 Week Highs Today "Kelly Short"
Previous Article (by Author): Re: [CANSLIM] Low Number... a bit OT Eric Jaenike
Top of Thread: Re: [CANSLIM] CANSLIM Hunting List 09/20/02 Spencer48@aol.com
Next in Thread: Re: [CANSLIM] CANSLIM Hunting List 09/20/02 Spencer48@aol.com
Articles sorted by: [Date]
[Author]
[Subject]
This is a multi-part message in MIME format. - ------=_NextPart_000_00D3_01C2649A.736A6950 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi jans and Patrick, I currently get the forward growth rate from VectorVest and haven't been = able to find any other sources that were as reliable for filtering = CANSLIM candidates. I know that DGO, Quotes Plus, and various internet = sites carry forward estimates, but VV takes it one step further. They do = a statistical analysis on past growth rates, looking for patterns that = would influence future estimates. While the actual formula is = proprietary, as I understand it, they "temper" the forward estimates = done by analysts if the volatility of past growth rates is fairly high = and/or slowing significantly.=20 I chose to filter stocks based on forward growth >=3D15 based on this = reasoning: I needed a way to identify growth stocks from the universe of = all possible stocks. By definition, growth stocks are those displaying = growth that exceeds the industry and/or market average. I didn't want to = rely on past growth rates because the market is a discounting mechanism, = i.e. discounting the future to measure "value" and then pricing = accordingly. That means it relies on future estimates more than past = results when pricing a stock. I then did a study of "average" forward = growth rates (in VV) for all stocks and found that, even at the market = peaks (where forecasts tend to overshoot significantly), the maximum = average forward growth rate was about 12-13%. Again, that's an *average* = of all stocks, so is comprised of stocks with extremely high and = extremely negative growth rates. I knew that my filter had to be higher = than the average. I then did some anecdotal backtesting, pulling up the = charts of many "big winner" stocks and looked at their pattern of = forward growth rates. As it turned out, >=3D15% turned out to be norm = for these stocks when they started their big moves. After that, I went = back in time and simply created subsets of stocks with forward growth = >=3D15% and sure enough, all the stocks that kept popping up in the IBD = were there.=20 With respect to whether or not these growth rates are useful because = they are prone to error....I don't worry about that. Forecasting is a = difficult process, but there is no way to operate in business or in the = stock market without it if you are trying to assess fundamental drivers = in some meaningful way. In business, the shorter the lead times for = acquiring capacity and material, the less far into the future you have = to "guess" about what the needs will be. But there's no way to operate = without some kind of forecasting. For example, if it takes 2 years to = bring a new plant on line, you have to know what the needs are going to = be 2 years into the future so that you can get the ball rolling. Will = the forecast be wrong? Often it is, but to ignore forecasts because they = are error prone leaves you in the dust when competitors have learned to = forecast with the least error possible and then move ahead as necessary = to ensure capacity is available to meet demand. The times where errors = are most prevalent are at the points in time where the trend changes. = So, if a market is expanding, you can make errors, but because the = upward trend is continuing, you can "make up" for the error over time. = However, when the market suddenly stops growing or worse, declines, the = forecasting errors become quite large. Take that same argument and apply it to forecasting future earnings for = a business for use in valuing stocks. I know that the estimates are = going to be wrong. That's the nature of forecasting. But how can I use = it, knowing that it is not a perfect predictor? Simple. Find a "line in = the sand" so to speak. I know, for example, when the trend is up (as it = was going in to the market peak in early 2000), the estimates are going = to be too high. Conversely, when the trend is down, at some point = estimates will be too *low*. Either way, it's not important to me. Why? = Because selecting based on an estimate of future growth is simply a = mechanism for *identifying* CANSLIM candidates. I can also look at the = *trend* in that forward growth rate for any particular stock. If the = trend has been steadily down and has begun to turn back up, I know = something is going on. If a business is suddenly beating estimates over = and over again, I know that the estimates are too low.=20 Will I miss some good quality stocks because of an arbitrary cut off at = 15%? Yes. But what I've found is that the few that I miss are a small = price to pay for the greater universe that I'm able to identify. Why use = forward growth rates based on estimates that may be wrong vs. past = growth rates that are based on actual financial reports? Because = fundamentals lag badly (I'm ignoring accounting fraud...that's another = ball of wax). I can easily find stocks that had 20-25% growth rates in = the past, but what if their business has changed? What if their market = has changed? What if new competitors have come in? What if alternative = products/services are now available? What if the economic backdrop has = changed? The future estimates at least keep that in mind. Once I find = stocks with forward growth rates "above average," then I can look at = technicals to see that the market is confirming the future potential of = the business, then I can look at past fundamentals to see how = effectively the company has generated profits in the past. All told, a = forward growth rate or a forward earnings estimate is just a means for = identifying stocks with potential. Unless I'm buying the whole company, = it doesn't really matter if it's "wrong." Katherine ----- Original Message -----=20 From: Patrick Wahl=20 To: canslim@lists.xmission.com=20 Sent: Tuesday, September 24, 2002 2:26 PM Subject: Re: [CANSLIM] CANSLIM Hunting List 09/20/02 On 24 Sep 2002 at 14:55, Spencer48@aol.com wrote: > about what the stock will do. But how do they divvy up the = quarterlies > (from their yearly projections)-and how many quarters ahead can they > accurately predict? Too often I think the correct answer is zero quarters. Right now I'd = say there is very=20 very little reliablity in any earnings forecast, because they are so = dependent on=20 when the economy turns up, and no one can make that call. During a = period of a=20 growing economy, forecasts are more reliable, its when you get into = these transition=20 periods that they completely lose their accuracy. No one was = forecasting a big slow=20 down in earnings as we approached the recent slowdown. Also, certain = industries=20 and business are easier to forecast for than others. Forecasts for = Abott Labs, when I=20 have paid attention to it, are right on the nose. I think smaller = companies,=20 especially in technology areas, are more difficult to forecast for. - -To subscribe/unsubscribe, email "majordomo@xmission.com" -In the email body, write "subscribe canslim" or -"unsubscribe canslim". Do not use quotes in your email. - ------=_NextPart_000_00D3_01C2649A.736A6950 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; = charset=3Diso-8859-1"> <META content=3D"MSHTML 6.00.2719.2200" name=3DGENERATOR> <STYLE></STYLE> </HEAD> <BODY bgColor=3D#ffffff> <DIV> <DIV>Hi jans and Patrick,</DIV> <DIV> </DIV> <DIV>I currently get the forward growth rate from VectorVest and haven't = been=20 able to find any other sources that were as reliable for filtering = CANSLIM=20 candidates. I know that DGO, Quotes Plus, and various internet sites = carry=20 forward estimates, but VV takes it one step further. They do a = statistical=20 analysis on past growth rates, looking for patterns that would influence = future=20 estimates. While the actual formula is proprietary, as I understand = it,=20 they "temper" the forward estimates done by analysts if the = volatility of=20 past growth rates is fairly high and/or slowing significantly. </DIV> <DIV> </DIV> <DIV>I chose to filter stocks based on forward growth >=3D15 based on = this=20 reasoning: I needed a way to identify growth stocks from the universe of = all=20 possible stocks. By definition, growth stocks are those displaying = growth=20 that exceeds the industry and/or market average. I didn't want to rely = on past=20 growth rates because the market is a discounting mechanism, i.e. = discounting the=20 future to measure "value" and then pricing accordingly. That means it = relies on=20 future estimates more than past results when pricing a stock. I then did = a study=20 of "average" forward growth rates (in VV) for all stocks and found = that,=20 even at the market peaks (where forecasts tend to overshoot = significantly), the=20 maximum average forward growth rate was about 12-13%. Again, that's an = *average*=20 of all stocks, so is comprised of stocks with extremely high and = extremely=20 negative growth rates. I knew that my filter had to be higher than the = average.=20 I then did some anecdotal backtesting, pulling up the charts of many = "big=20 winner" stocks and looked at their pattern of forward growth rates. As = it turned=20 out, >=3D15% turned out to be norm for these stocks when they started = their big=20 moves. After that, I went back in time and simply created subsets of = stocks with=20 forward growth >=3D15% and sure enough, all the stocks that kept = popping up in=20 the IBD were there. </DIV> <DIV> </DIV> <DIV>With respect to whether or not these growth rates are useful = because they=20 are prone to error....I don't worry about that. Forecasting is a = difficult=20 process, but there is no way to operate in business or in the stock = market=20 without it if you are trying to assess fundamental drivers in some = meaningful=20 way. In business, the shorter the lead times for acquiring capacity and=20 material, the less far into the future you have to "guess" about what = the needs=20 will be. But there's no way to operate without some kind of forecasting. = For=20 example, if it takes 2 years to bring a new plant on line, you = have to=20 know what the needs are going to be 2 years into the future so that you = can get=20 the ball rolling. Will the forecast be wrong? Often it is, but to ignore = forecasts because they are error prone leaves you in the dust when = competitors=20 have learned to forecast with the least error possible and then move = ahead as=20 necessary to ensure capacity is available to meet demand. The times = where errors=20 are most prevalent are at the points in time where the trend changes. = So, if a=20 market is expanding, you can make errors, but because the upward trend = is=20 continuing, you can "make up" for the error over time. However, when the = market=20 suddenly stops growing or worse, declines, the forecasting errors become = quite=20 large.</DIV> <DIV> </DIV> <DIV>Take that same argument and apply it to forecasting future earnings = for a=20 business for use in valuing stocks. I know that the estimates are going = to be=20 wrong. That's the nature of forecasting. But how can I use it, knowing = that it=20 is not a perfect predictor? Simple. Find a "line in the sand" so to = speak. I=20 know, for example, when the trend is up (as it was going in to the = market peak=20 in early 2000), the estimates are going to be too high. Conversely, when = the=20 trend is down, at some point estimates will be too *low*. Either way, = it's not=20 important to me. Why? Because selecting based on an estimate of future = growth is=20 simply a mechanism for *identifying* CANSLIM candidates. I can also look = at the=20 *trend* in that forward growth rate for any particular stock. If the = trend has=20 been steadily down and has begun to turn back up, I know something is = going on.=20 If a business is suddenly beating estimates over and over again, I know = that the=20 estimates are too low. </DIV> <DIV> </DIV> <DIV>Will I miss some good quality stocks because of an arbitrary cut = off at=20 15%? Yes. But what I've found is that the few that I miss are a small = price to=20 pay for the greater universe that I'm able to identify. Why use forward = growth=20 rates based on estimates that may be wrong vs. past growth rates that = are based=20 on actual financial reports? Because fundamentals lag badly (I'm = ignoring=20 accounting fraud...that's another ball of wax). I can easily find stocks = that=20 had 20-25% growth rates in the past, but what if their business has = changed?=20 What if their market has changed? What if new competitors have come in? = What if=20 alternative products/services are now available? What if the=20 economic backdrop has changed? The future estimates at least = keep that=20 in mind. Once I find stocks with forward growth rates "above average," = then I=20 can look at technicals to see that the market is confirming the future = potential=20 of the business, then I can look at past fundamentals to see how = effectively the=20 company has generated profits in the past. All told, a forward growth = rate or a=20 forward earnings estimate is just a means for identifying stocks with = potential.=20 Unless I'm buying the whole company, it doesn't really matter if it's=20 "wrong."</DIV> <DIV> </DIV> <DIV>Katherine</DIV></DIV> <BLOCKQUOTE=20 style=3D"PADDING-RIGHT: 0px; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; = BORDER-LEFT: #000000 2px solid; MARGIN-RIGHT: 0px"> <DIV style=3D"FONT: 10pt arial">----- Original Message ----- </DIV> <DIV=20 style=3D"BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: = black"><B>From:</B>=20 <A title=3Dpjwahl@attbi.com href=3D"mailto:pjwahl@attbi.com">Patrick = Wahl</A>=20 </DIV> <DIV style=3D"FONT: 10pt arial"><B>To:</B> <A = title=3Dcanslim@lists.xmission.com=20 = href=3D"mailto:canslim@lists.xmission.com">canslim@lists.xmission.com</A>= </DIV> <DIV style=3D"FONT: 10pt arial"><B>Sent:</B> Tuesday, September 24, = 2002 2:26=20 PM</DIV> <DIV style=3D"FONT: 10pt arial"><B>Subject:</B> Re: [CANSLIM] CANSLIM = Hunting=20 List 09/20/02</DIV> <DIV><BR></DIV><BR><BR>On 24 Sep 2002 at 14:55, <A=20 href=3D"mailto:Spencer48@aol.com">Spencer48@aol.com</A> = wrote:<BR><BR>> about=20 what the stock will do. But how do they divvy up the = quarterlies<BR>>=20 (from their yearly projections)-and how many quarters ahead can = they<BR>>=20 accurately predict?<BR><BR>Too often I think the correct answer is = zero=20 quarters. Right now I'd say there is very <BR>very little = reliablity in=20 any earnings forecast, because they are so dependent on <BR>when the = economy=20 turns up, and no one can make that call. During a period of a=20 <BR>growing economy, forecasts are more reliable, its when you get = into these=20 transition <BR>periods that they completely lose their accuracy. = No one=20 was forecasting a big slow <BR>down in earnings as we approached the = recent=20 slowdown. Also, certain industries <BR>and business are easier = to=20 forecast for than others. Forecasts for Abott Labs, when I <BR>have = paid=20 attention to it, are right on the nose. I think smaller = companies,=20 <BR>especially in technology areas, are more difficult to forecast=20 for.<BR><BR><BR><BR>-<BR>-To subscribe/unsubscribe, email "<A=20 = href=3D"mailto:majordomo@xmission.com">majordomo@xmission.com</A>"<BR>-In= the=20 email body, write "subscribe canslim" or<BR>-"unsubscribe = canslim". Do=20 not use quotes in your email.</BLOCKQUOTE></BODY></HTML> - ------=_NextPart_000_00D3_01C2649A.736A6950-- - - - -To subscribe/unsubscribe, email "majordomo@xmission.com" - -In the email body, write "subscribe canslim" or - -"unsubscribe canslim". Do not use quotes in your email.
Next Article (by Author): [CANSLIM] Low Number of New 52 Week Highs Today "Kelly Short"
Previous Article (by Author): Re: [CANSLIM] Low Number... a bit OT Eric Jaenike
Top of Thread: Re: [CANSLIM] CANSLIM Hunting List 09/20/02 Spencer48@aol.com
Next in Thread: Re: [CANSLIM] CANSLIM Hunting List 09/20/02 Spencer48@aol.com
Articles sorted by: [Date]
[Author]
[Subject]