How do you set a price target?

aachimp's picture
Rank: Gorilla | 662

hey guys, I'm learning about stock valuation and was wonderin if you someone knew how to set a convincing price target. what types of analysis would you have to use to go about it? thanks

Comments (46)

Mar 17, 2007

I think most people do a DCF analysis. And then whatever you come up with as the equity value you then use as the target price. Its hard to be anymore specific than that unless you want to clarify your question further.

Mar 18, 2007

Forecast next years EPS using financial modelling based on company and market knowledge. Determine appropriate market multiple based on industry comparables and historical valuations. EPS x Multiple = Target price.

Mar 19, 2007
iceman:

Forecast next years EPS using financial modelling based on company and market knowledge. Determine appropriate market multiple based on industry comparables and historical valuations. EPS x Multiple = Target price.

Could add in EV/Sales and EV/EBITDA industry multiples along with P/E. The more robust and inclusive your models, the better the price target.

Mar 18, 2007

I've heard there is a really advanced magic eight ball (MEB) analysis that you can do.

Mar 20, 2007

MEB is very difficult analysis as the EB tends to only accept Binary questions. If your not careful, youll have the answer to wrong question.

Mar 9, 2008
aachimp:

hey guys, I'm learning about stock valuation and was wonderin if you someone knew how to set a convincing price target. what types of analysis would you have to use to go about it? thanks

Really depends. You can use all sorts of multiples such as FCF, EBIT, EPS, ROE, etc to come up with a target price.

Mar 9, 2008

1) Open Excel and import 1-yr of historical EOD price data for that company
2) Calculate Mean chg per day
3) Calculate Standard Deviation
4) Set up 120 rows to represent trading days... day 0 = current price.
5) On day 1, multiply day 0 cell by (1 + Norminv(Rand(),mean cell, std cell))... don't forget to pray to the normal distribution gods
6) On day 2, do the same as day 1... drag to repeat for all trading days. Make sure it compounds the % chg each day.

Repeat simulation and cherish that normal distribution. Don't think about tail scenarios. And mathematically, your safest target price is today's closing price. I apologize for my cynicism.

More seriously, google DCF analysis if you think you can predict future CFs (or use historical earnings and make assumption of trend provided a stable business).. No guarantee that it will trade near that price but that should be the best "intrinsic" value data point.

Jul 12, 2007

I'll tell you the answer if you really want to know...

...but I warn you that it won't be what you are expecting.

Mar 9, 2008

Always one year ahead from last reporting period. You may have to use a blend of two years to get your TP, for example - 50% FY16 and 50% FY17.

Mar 9, 2008

Price targets are generally the analyst projections of a stock's future price for a given period of time. They are definitely helpful to see how you own valuation holds up.

Mar 9, 2008

at my firm (BB), it's 1 year from whenever that report was written.

Mar 9, 2008

Ah I see - so if the report was written on June 1, 2015, the 1 year price target would be for June 1, 2016, is that right?

Also along those lines, if that report was written on June 1, 2015 (and the company's FY is simply from Jan-Dec), presumably the LTM earnings would have been used, and so if they used a DCF, the DCF would be based off of LTM figures and not the last FY numbers, is that right?

Mar 9, 2008

correct

Mar 9, 2008

Great thank you all!

Jul 12, 2007

Two ways and then trying to reconcile between them:

1) relative valuation - comparing to other companies using tons of ratios as tools
2) DCF - discounting cash flows by finding WACC or Return on Equity

During training they will explain when one is inappropriate and when one is preferred (ex. company experiencing losses, company is high growth, etc.).

Or listen to inscrutable, he actually is a equity researcher whereas I'm a summer analyst.

Mar 9, 2008

one more thing. i want to get the 1 year target price.

Mar 9, 2008

Any type of comp based valuation (transaction or comperable) is, like the B / S, taken at a moment in time. I.e. you're not forecasting future transactions or forward market data. So any multiple you arrive at is as of the day you compile the data.

That said, simply multiply your 2014 multiple by your EV/Sales 2011E.

Mar 9, 2008
Stringer Bell:

Any type of comp based valuation (transaction or comperable) is, like the B / S, taken at a moment in time. I.e. you're not forecasting future transactions or forward market data. So any multiple you arrive at is as of the day you compile the data.

That said, simply multiply your 2014 multiple by your EV/Sales 2011E.

i'm assuming you mean 2014 sales X EV/Sales for 2011? so that would give me a EV, but how do I get the 1 year target price from that?

Mar 9, 2008

.

Mar 9, 2008

There is something missing here. I remember some sort of discounting somewhere.

Mar 9, 2008

What has been discussed so far is really more of what you think fair value is for the stock today. I.e. if you think it should trade at parody with its comp group, you would apply 6x multiple (the peer multiple) to your 2014E sales and determine the equity value per share today (or tomorrow as indicated).

What it "should" trade at today can be very different from your price target methodology. For your 1 year price target, you can set a valuation which the stock will need to grow into. For instance, let's say the comp group trades for 8x EV/Sales 24 months out. Then you could take an 8x multiple to your 2014E estimate (effectively 36 months out) and suggest this will be the EV in 12 months. Also, in this case you need to forecast the cap structure to 1 year out and find the new derived equity value per share.

Personally, I think the first method is not a price target at all, but it surprises me how many guys on the street use fair value and price target interchangeably.

Mar 9, 2008

I should have also mentioned that comp'ing to a peer multiple in 2014E will not have much integrity because the number of analysts publishing 2014E numbers - even in a heavily covered name - is likely very thin.

To your point, this is where discounting comes into play. You could use a 2012E multiple applied to your 2014E sales and then discount the results. The discount rate will be company specific.

Mar 9, 2008

jb that helps a lot. appreciate it.

Mar 9, 2008

and you are totally right when you say that people use fair value and price target interchangeably...it pisses me off.

Jul 12, 2007

ThoughtMan has the textbook answer - although I would throw in a third one which is breakup/takeout/scrap value (although technically this is a derivative of number 2 - as is EVA).

The true answer, however, lies in your question - one word in particular in the first sentence.

Mar 9, 2008

do your own HW

I eat success for breakfast...with skim milk

    • 1
Mar 9, 2008
TonyPerkis:

do your own HW

I've been done with "HW" for years.

    • 2
Mar 9, 2008

Not an expert at valuation by any means. But some thoughts:

Did you remember to subtract out net debt when you are using EV/Sales as comp ratio?
Typically, more than one ratio would be used to find a target range.
How conservative are your revenue projections?

Mar 9, 2008

What is the problem?

Mar 9, 2008

Not sure what the problem is.. your description was fairly vague... but your sum of parts valuation should be yielding something fairly close to the current price.. assuming your mechanics are correct, i.e. share value calc from EV as suggested above...

its really no different than taking a weighted average multiple and using that to value the entire firm...

check this: is the current trading multiple significantly different than the comps being used? i.e. is your company trading at 20x EBITDA when the comps (in both different industries) trading at 8x-10x?? There is often a reason that comps don't align, may be growth prospects? but if your company is currently trading at a multiple outside of it's range then you may need to adjust for the differentiating factor in order to arrive at a "true" fair value.

Mar 9, 2008

Depending on the industry EV/Sales may not be a good metric. For example, if the industry is composed of competitors with different business models (high margin/low volume and high volume/low margin) then averaging the EV/Sales of those competitors is going to give you a bogus comp. Do some research to see if EV/Sales is a common valuation tool in the industries you are comparing. Other thoughts:
- Like another commenter said, make sure you subtract net debt from EV to get to your equity value when comparing stock price
- Sum of parts may not be appropriate depending on the industries the businesses engaged in. If the businesses are highly correlated investors should require a greater return on equity (lower stock price) to own them.

Mar 9, 2008
Boothorbust:

Depending on the industry EV/Sales may not be a good metric. For example, if the industry is composed of competitors with different business models (high margin/low volume and high volume/low margin) then averaging the EV/Sales of those competitors is going to give you a bogus comp. Do some research to see if EV/Sales is a common valuation tool in the industries you are comparing. Other thoughts:
- Like another commenter said, make sure you subtract net debt from EV to get to your equity value when comparing stock price
- Sum of parts may not be appropriate depending on the industries the businesses engaged in. If the businesses are highly correlated investors should require a greater return on equity (lower stock price) to own them.

EV/EBITDA is the common metric, but I decided to use EV/S because most of its competitors are foreign and accounting practices can "fudge" EBITDA numbers with foreign firms.

Mar 9, 2008

Just some quick notes.

There is no debt on its books so no need to subtract it out.

The firm's EV/S is in line with its competitors and customers in industry A. However, in the firm's industry B, the EV/S is half of industry average. So basically I took the revenue from the firm's one industry and took the industry average and valued its accordingly. I think this is where my target price is getting messed up, however I think it's appropriate to do a sum of parts valuation because the firm's industries are very different. If I use the EV/S from industry A on the revenues for industry B. The price target becomes more realistic and is definitely in line with the firm's current share price.

Thoughts?

Mar 9, 2008

so it issssssssss hwwww

I eat success for breakfast...with skim milk

Mar 9, 2008
TonyPerkis:

so it issssssssss hwwww

Another added value user. Why does this forum reward useless users such as yourself?

Jul 14, 2007

also, in research, talking with company execs may help.

disclaimer: not in er

Jul 14, 2007

Convincing is relative - it's bullshit in and bullshit out to be honest - you can convince anyone of any real particularly robust number that is ahead of consensus by adding all sorts of flairs.

Depends on the sector as well - DCF is horseshit unless you're in pharma or a few others, but a price target takes into consensus - where consensus is wrong and what you're taking into account and what they are not.

Jul 14, 2007

I see people using relative valuation to find a reasonable upper and lower bounds, and using DCF to back it up (ie confirming number from relative valuation.)

Jul 14, 2007
Comment
Oct 28, 2007