Help needed with EV/EBITDA

Hi all.

I need help with a formula I’m trying to implement in a Ranking.

I want to place in the ranking the formula EV/EBITDA, but I doubt about the correct way to implement it, and what’s the best factor for EBITDA.

For EBITDA I am considering the following factors:

A- EBITDMgn%TTM
B- OpIncBDeprTTM
C- EBITDAPTM

But also I’m not sure about what’s best, TTM or PTM?

And, how should I write the formula?

EV/ EBITDA or EBITDA/EV?

And the arrow… up or down?

Thanks!!!


Captura de pantalla 2019-11-06 a las 11.49.11.png

This is the formula that I have been using. If there’s a tweak that someone wants to make to it them I’m happy to accept suggestions.

EV/EBITDA TTM = 1/((MktCap - CashPSQ * ShsOutMR)/Eval(EBITDATTM>0,EBITDATTM,NA))
Lower

Hi Monday, I use the yield version EBITDA TTM / EV. Don’t know it’s the best way, but it’s what I use.

EBITDA(0,TTM) / EV

Higher is better, lower is worse.

Barn et all - you should put EV in the denominator as this factor has the lower probability of being negative. The formula I would use is as follows:

EBITDATTM/abs(EV)

Higher values are better from a valuation perspective thus the little arrow should point up. It is necessary to use the absolute value of EV because there is a slight chance that EV could be negative. For ranking purposes, it is undesirable to have a negative value in the denominator and although using abs(EV) is not correct, it is the lesser of two evils.

Also, there is no reason to provide a formulation from first principles. Both EV and EBITDATTM are provided by P123.

Steve

WOW!!, thanks a lot for your help guys, you are awesome!!

Just a note: EV has a zero chance of being negative. If it’s negative then we return NA.

Great, thanks Paul.

I’m not sure why P123 would do that. A negative EV is legitimate and there are proponents out there who suggest that negative EV companies perfom quite well.

Steve

I humbly suggest that P123 revisit negative EVs. It is not an impossible situation and should not be treated as such.

https://blogs.cfainstitute.org/investor/2013/07/10/returns-on-negative-enterprise-value-stocks-money-for-nothing/

Steve

A negative EV generally means that the company has a massive cash position. The original intention of the EV was to come up with a number that an acquirer might use to value the company. But as much as we might like the idea, no one will pay you money to take all that money off their hands.

And from a more practical perspective, the most common use of EV in my experience is EV/EBITDATTM (or the inverse). If EV is negative then you’re in special-condition territory for valuation purposes. You would favor negative EBITDA, the more negative the better. It’s generally easier to set EV to NA to eliminate those situations and don’t worry about it.

You’re welcome to calculate EV from scratch and retain all the negatives.

Practical for whom?

If you are using the factor in a ranking system, the worst thing you can do is put EBITDA in the denominator, assuming P123 isn’t N/A 'ing negative earnings. Do you really want negative earnings to outrank positive earnings? Seriously???

True - but the average investor that may arrive at P123 isn’t aware of the “big brother” initiatives in place here. P123 should be implementing formulae as they are supposed to be, not what they believe is in the best interests of users of the site, which in many instances, isn’t.

Paul, I appreciate your contributions here, but I have to disagree on this one.

My first thought was to suggest to make sure it is mentioned in the Factor Reference.

I just checked and -fair point to P123- it is already there…

"Also, it is possible for an extremely cash-heavy company to have a negative enterprise value (i.e. the market is, for one reason or another, not paying attention to the full amount of cash). In the latter case, we show Enterprise Value and Enterprise Value per share as NA. "

Anyways, I am going to reiterate what I believe you guys should be doing.

First of all, use the formula EBITDATTM/Abs(EV). Keep the Abs() even though it is not necessary with the P123 implementation of EV. When you move to any other platform (such as Quantopian) the Abs() will be needed. So you might as well keep it as a reminder.

Second, a good rule of thumb is to keep the factor that can be negative or positive in the numerator, never put it in the denominator as it is problematic. In the case of EBITDA being on the bottom, you would need to select “lower is better” (little arrow down) and companies with negative earnings would then be ranked higher than companies with positive earnings, right?. After all, negative is lower than positive. By putting EBITDA in the numerator on the other hand, you then need to select “higher is better” for the ranking algorithm (little arrow up). In this case, positive earnings ALWAYS rank higher than negative earnings. Most ranking system designers believe that positve earnings are better than negative earnings for a value factor. However, if you are not one of them then sure… put the earnings in the denominator.

Take care all,
Steve