sharetrader
Page 1 of 2 12 LastLast
Results 1 to 10 of 15
  1. #1
    always learning ... BlackPeter's Avatar
    Join Date
    Aug 2007
    Posts
    9,497

    Default How good are the forecasts of stockmarket analysts?

    We love to be cynical and beat stock market analysts up while secretly still following their analysis, at least if it suits our agenda - don't we?

    I thought I try to put a bit of analysis into this area, then at least we will know.

    Rules of the game:

    I shall take the consensus and the recommendation of NZX listed companies which are regularly analysed by more than one analyst and compare 12 months later what really happened. Companies researched will come off my watch list (in this case I have all the numbers anyway available).

    Consensus and recommendation taken from marketscreener.com

    If the share price of the stock in forecast month is within consensus +/- 10% then I call that a PASS, otherwise I call that a FAIL. Fair enough?

    If recommendation is HOLD, I call it a pass if stock performed similar to NZX50 +/- 10%
    If recommendation is OUTPERFORM or BUY I call it a pass if stock outperformed NZX50
    If recommendation is UNDERPERFORM or SELL I call it a pass if stock underperformed NZX50.

    These are the companies and forecasts I looked at so far ...

    Stock prediction timeframe analyst hitrate link to post
    ABA Feb-19 to FEb-20 Shareprice Forecast Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showth...l=1#post796127
    AIA Feb-19 to Feb-20 Shareprice Forecast Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showth...l=1#post801656
    ANZ Jan-19 to Jan-20 Shareprice Forecast Pass; Buy Recommendation: Fail; http://://www.sharetrader.co.nz/show...ANZ#post791775
    A2M / ATM Jan-19 to Jan-20 Shareprice Forecast: Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?2318-ATM-A2-Corporation-Limited&p=792116&viewfull=1#post792116
    FBU Feb-19 to Feb-20 Shareprice Forecast: Pass; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?10702-Fletcher-Building-(FBU)&p=795413&highlight=FBU#post795413
    FPH Jan-19 to Jan-20 Shareprice Forecast: Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?8537-Fph&p=791601&viewfull=1#post791601
    HGH Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showthread.php?8425-HGH-Heartland-Group-Holdings&p=791874&viewfull=1#post791874
    KMD Feb-19 to Feb-20 Shareprice Forecast Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showthread.php?7128-Kathmandu-IPO&p=797912#post797912
    MCY Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?9340-Meridian-Float&p=791423&viewfull=1#post791423
    MFT Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showthread.php?760-MFT&p=791142&viewfull=1#post791142
    MPG Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?9836-Metro-Glass-IPO&p=792449&viewfull=1#post792449
    OCA Feb-19 to Feb-20 Shareprice Forecast Pass; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?9856-Oceania-Group-retirement-villages&p=799799&highlight=OCA#post799799
    SCL Feb-19 to Feb-20 Shareprice Forecast Pass; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showth...135#post796135
    SPK Feb-19 to Feb-20 Shareprice Forecast Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showth...SPK#post798675
    STU Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: PASS; https://www.sharetrader.co.nz/showthread.php?1835-STU&p=792867&viewfull=1#post792867
    SUM Jan-19 to Jan-20 Shareprice Forecast Fail; Buy Recommendation: FAIL; https://www.sharetrader.co.nz/showthread.php?5009-Summerset-Group-IPO&p=792252&viewfull=1#post792252
    THL Jan-19 to Jan-20 Shareprice Forecast: Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?5278-THL-Tourism-Holdings&p=790937&viewfull=1#post790937
    TPW Feb-19 to Feb-20 Shareprice Forecast: Fail; Buy Recommendation: Pass; https://www.sharetrader.co.nz/showthread.php?1363-TRUST-POWER&p=796435&viewfull=1#post796435
    XRO Feb-19 to Feb-20 Shareprice Forecast: Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?4774-Xro-xero&highlight=XRO&p=794962#post794962
    ZEL Feb-19 to Feb-20 Shareprice Forecast: Fail; Buy Recommendation: Fail; https://www.sharetrader.co.nz/showthread.php?9253-Z-Energy-IPO&p=800689&highlight=ZEL#post800689

    Rating of the analyst consensus forecasts so far:

    20 stock forecasts checked so far (checking for each consensus and buy recommendation);
    Consensus share price forecasts correct: 4/20; analysts hitrate: 20%
    Consensus recommendation vs NZX50 correct: 7/20; analysts hitrate: 35%

    Work in progress - bear with me .
    Last edited by BlackPeter; 23-03-2020 at 03:23 PM. Reason: work in progress ...
    ----
    "Prediction is very difficult, especially about the future" (Niels Bohr)

  2. #2
    On the doghouse
    Join Date
    Jun 2004
    Location
    , , New Zealand.
    Posts
    9,222

    Default

    Quote Originally Posted by BlackPeter View Post
    Analyst hit rate so far: .
    Hi BP, interesting exercise. I am a little confused by your 'analyst hitrate' ratio though, and I will use ANZ as an example. You say it is 1:1 (?)

    You state in the ANZ thread that you are looking at the guesses of 15 analysts. From those you derive a 'consensus forecast' of $28.68. Is that just an average of the 15 analysts guesses?

    I will give you my understanding of how your rules work in the case of ANZ

    1/ First select a baseline. You start with the 'peak price in the prediction month' of January 2019 of $26.19.
    2/ The consensus forecast of $28.68 is 9.5% above that. So you are calling this prediction a 'hold' as the gain predicted is under 10%
    3/ In order for your 'analyst consensus' to be correct that means the actual gain must be within 10% of the start price. The peak price in January 2020 was $25.90. $25.90 is a loss of 1% from the base value. This is within the plus or minus 10% target you set for a hold recommendation. So in this case you regard the consensus of analysts views as 'satisfactory'.

    Now please tell me how that result (if I have gone through all the steps correctly) translates to that 1:1 ratio figure in your table!

    SNOOPY
    Last edited by Snoopy; 15-02-2020 at 02:56 PM.
    Watch out for the most persistent and dangerous version of Covid-19: B.S.24/7

  3. #3
    always learning ... BlackPeter's Avatar
    Join Date
    Aug 2007
    Posts
    9,497

    Default

    Quote Originally Posted by Snoopy View Post
    Hi BP, interesting exercise. I am a little confused by your 'analyst hitrate' ratio though, and I will use ANZ as an example. You say it is 1:1 (?)

    You state in the ANZ thread that you are looking at the guesses of 15 analysts. From those you derive a 'consensus forecast' of $28.68. Is that just an average of the 15 analysts guesses?

    I will give you my understanding of how your rules work in the case of ANZ

    1/ First select a baseline. You start with the 'peak price in the prediction month' of January 2019 of $26.19.
    2/ The consensus forecast of $28.68 is 9.5% above that. So you are calling this prediction a 'hold' as the gain predicted is under 10%
    3/ In order for your 'analyst consensus' to be correct that means the actual gain must be within 10% of the start price. The peak price in January 2020 was $25.90. $25.90 is a loss of 1% from the base value. This is within the plus or minus 10% target you set for a hold recommendation. So in this case you regard the consensus of analysts views as 'satisfactory'.

    Now please tell me how that result (if I have gone through all the steps correctly) translates to that 1:1 ratio figure in your table!

    SNOOPY
    Not quite, but I agree it is complicated ;

    1) Well - yes, the "consensus" is (I think) the median of all known analyst forecasts. Many web pages calculate that for you.

    If after 12 months the share price hits this consensus (+/- 10%), then I give the analysts one point, if it further away I don't.

    The other point comes from the analyst recommendation (again median) - e.g. "BUY", "HOLD", "SELL" which the analysts produce independently from their consensus.

    I check what the share price did over the last 12 months compared to the NZX 50. If it moved in line with the NZX50, than I think a HOLD recommendation was appropriate (i.e. another point for the analysts). Same if it outperformed the NZX50 and the recommendation was Öutperform"or BUY"or it it underperformed the NZX50 and the recommendation ws underperform or sell.

    If however the analysts recommend a SELL or Underperform (as with e.g. MFT), but the Share price developed 50% better than the NZX50 (which did amazing as well), than I say the analyst recommendation was not helpful. No point.

    In the case of ANZ: Analysts predicted a share price of A$28.68 for January 2020, and it reached A$25.90. Not flash, but less than 10% below the prediction, so I called that "'pass"

    However - analysts said as well in January 2019 that ANZ will "outperform" other stocks. NZX went up over the next 12 months by something like 27%, while the ANZ share price was after 12 months the same as before. It did not outperform the stock markets, i.e. I call this a fail. This is the 1:1;

    Does this help?
    ----
    "Prediction is very difficult, especially about the future" (Niels Bohr)

  4. #4
    On the doghouse
    Join Date
    Jun 2004
    Location
    , , New Zealand.
    Posts
    9,222

    Default

    Quote Originally Posted by BlackPeter View Post
    Does this help?
    Ah so you actually doing two separate tests, which are nevertheless slightly different ways of looking at the same thing (how good brokers are with their picks). Your 'ratio' is actually two columns, the first denoting a satisfactory result and the second being unsatisfactory.

    With your result scoring system there are three possible outcomes:

    1/ Thus the result you are hoping for is '2:0' , a pass for both tests.
    2/ '1:1' indicates one pass and one fail AND
    3/ 0:2' indicates a double fail

    Got it.

    SNOOPY
    Last edited by Snoopy; 15-02-2020 at 04:41 PM.
    Watch out for the most persistent and dangerous version of Covid-19: B.S.24/7

  5. #5
    Speedy Az winner69's Avatar
    Join Date
    Jun 2001
    Location
    , , .
    Posts
    37,738

    Default

    I wonder how good the analyst earnings forecasts are?

    They might be pretty good at doing the financial forecast but useless at assessing the market sentiment.

    Like they could be spot on EPS and their target price is based on a sensible logical assessment ....but the market goes gangbusters and share prices go up to ridiculous levels ....and their consensus targets are seen as pretty useless.

    Maybe you are just showing how over valued the NZX is at the moment.
    “ At the top of every bubble, everyone is convinced it's not yet a bubble.”

  6. #6
    Senior Member
    Join Date
    Nov 2018
    Location
    Christchurch
    Posts
    1,063

    Default

    Just buy the index and forget about the advisors. One would do a lot better to know in bad times (and it's only a matter of time when a crash happens), to have cash ready on the side to invest. THIS no financial advisor is going to tell you to do this... but they certainly know how to sell on you!

    If any manged fund or advisor produces a return or price that is THE SAME as something like the NZ50 - then they have STILL failed their job because the investor is still billed with their admin / mgt fee.

    In other posts i've mentioned the issue with such managed funds and those that give financial advice. Specifically links around Warren Buffet's view and his 10 year bet against managed funds how they would perform worse than the S&P500 index return. I would not think the NZX50 would be any different.

  7. #7
    always learning ... BlackPeter's Avatar
    Join Date
    Aug 2007
    Posts
    9,497

    Default

    Quote Originally Posted by Snoopy View Post
    Ah so you actually doing two separate tests, which are nevertheless slightly different ways of looking at the same thing (how good brokers are with their picks). Your 'ratio' is actually two columns, the first denoting a satisfactory result and the second being unsatisfactory.

    With your result scoring system there are three possible outcomes:

    1/ Thus the result you are hoping for is '2:0' , a pass for both tests.
    2/ '1:1' indicates one pass and one fail AND
    3/ 0:2' indicates a double fail

    Got it.

    SNOOPY
    Correct. I realize now as well that I could make that a bit more transparent. Will think about separating these two tests better in the results.
    ----
    "Prediction is very difficult, especially about the future" (Niels Bohr)

  8. #8
    always learning ... BlackPeter's Avatar
    Join Date
    Aug 2007
    Posts
    9,497

    Default

    Quote Originally Posted by winner69 View Post
    I wonder how good the analyst earnings forecasts are?

    They might be pretty good at doing the financial forecast but useless at assessing the market sentiment.

    Like they could be spot on EPS and their target price is based on a sensible logical assessment ....but the market goes gangbusters and share prices go up to ridiculous levels ....and their consensus targets are seen as pretty useless.

    Maybe you are just showing how over valued the NZX is at the moment.
    You might be right, though some of the stocks so far did ways outperform analyst forecasts and others did underperform them ...

    As well, while you are absolutely correct that share price is value plus hype, am I not sure how to separate these two components. Not sure either whether it would be any helpful if you want to use analyst predictions to make money (I suppose most of us do). Nobody pays you for value without the hype.

    Given that "the market is always right" is this as well the only fair method to assess the quality of forecasts.
    ----
    "Prediction is very difficult, especially about the future" (Niels Bohr)

  9. #9
    Speedy Az winner69's Avatar
    Join Date
    Jun 2001
    Location
    , , .
    Posts
    37,738

    Default

    Quote Originally Posted by BlackPeter View Post
    You might be right, though some of the stocks so far did ways outperform analyst forecasts and others did underperform them ...

    As well, while you are absolutely correct that share price is value plus hype, am I not sure how to separate these two components. Not sure either whether it would be any helpful if you want to use analyst predictions to make money (I suppose most of us do). Nobody pays you for value without the hype.

    Given that "the market is always right" is this as well the only fair method to assess the quality of forecasts.
    Like it ....’the market is always right’ and your work seems to be heading towards ‘analysts are generally wrong’

    But as long as punters believe they’re gurus they’ll still OK
    Last edited by winner69; 16-02-2020 at 10:49 AM.
    “ At the top of every bubble, everyone is convinced it's not yet a bubble.”

  10. #10
    always learning ... BlackPeter's Avatar
    Join Date
    Aug 2007
    Posts
    9,497

    Default

    Just in case anybody is wondering why I didn't continue with this analysis for March ...

    Well, while it looks like so far that the analyst predictions are pretty random anyway ... no analyst I am aware of could predict this virus in March 2019, i.e. it would be pretty meaningless for the current bear / high volatility phase to check their one year old stock price forecasts against what really happened during this "black swan event".

    I might continue this analysis starting again in April 2021 (and comparing forecasts against April 2020 numbers).

    For what it is worth ... I looked so far into 20 companies (with forecasts made in January and February 2019) vs outcomes in January and February 2020.

    Instead of just using my (obviously quite arbitrary) pass / fail criteria as per the top post - here is a scatter graph showing the forecasted SP change vs the observed SP change.

    Scattergraph analystpredictions.JPG

    As we can see: 20 consensus predictions, 5 of them too optimistic (forecasted consensus 10% or more above achieved SP), 4 of them "realistic (consensus matched target +/-10%), and 11 too pessimistic (target more than 10% above consensus).

    Slightly surprising, since we always seem to think that analysts are too optimistic, but they probably just didn't expect the last bull to run as far as it did.

    Whatever it is - based on above analysis it appears that throwing darts while being blindfolded might be an absolutely adequate method compared to using analyst consensus to determine where share prices for individual stocks will go.
    ----
    "Prediction is very difficult, especially about the future" (Niels Bohr)

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •