Hey guys, this may sound odd, but I have an Equity Research interview coming up and the interviewer has told me to expect McK and BCG Case interview type questions. I think its most likely going to be market sizing questions although I will prepare for other types of questions too.
Where should I go to best understand how to best approach market sizing questions ? I've only had one case interview before and I screwed up the market sizing question, so I want to have a better framework this time.
Thanks
Wetfeet Ace the Case.
Not odd at all. I interviewed for both ER and Consulting positions, and sizing questions came up often in ER. I also got asked valuation questions in a case interview style, i.e. structuring your approach, rather than describing various metrics, so that would be worth preparing for.
The McK, BCG, Bain, OC&C, OW websites all have sample questions and walk-throughs that would be worth looking at. From memory, McK and Bain are good for market sizing. However, in an ER interview the questions will be much shorter and simpler as less interview time is dedicated to them, so you can't afford to be as detailed as you would in a Consulting one.
For examples: GS ER: I was asked to estimate the number of planes flying over the UK at a given time. JPM ER: Estimating the number of letters delivered every day by the mail service. Both are very brief, compared to... OW: Estimating the market size of a firm selling baby clothes through a franchise store model, entering a new market.
Thanks a lot Porcine.
Look up examples of these questions online. Once you get an idea of how to approach them they become much easier.
Finished all the Wetfeet and Wharton examples. Anyone have links to other examples ?
If you type out your answers to the ones Porcine gave, people can critique you. Otherwise, google around a bit, there are definitely case books out there with market sizing questions.
Also: How many fathers are there in the US? How many post offices in the US? How many mailboxes? What's the market size for cars in the US? Market size for golf balls? Gillette razor blades?
How big of a parking lot would you need to fit every car in the world?
Estimate the market value of all real estate in Manhattan... I got this one in a McK interview and would like to see what solutions you guys suggest.
I don't know anything about Manhattan (been to NY only once) so my assumptions are probably way off, but this would be my approach:
Assumptions & Structure Guesstimate Manhattan = ~12k acres since I have no idea; ask interviewer if that's a fair assumption.
From what I know about Manhattan, its a business / downtown city type of area; assume there are only skyscrapers (30+ floors) there.
I know normal suburban plots (like my house) are about 1.5 acres, and just as a guesstimate it'd take about 4 plots of that size to fit a skyscraper, so 1 skyscraper for every 6 acres = ~2000 skyscrapers in Manhattan.
Breakdown skyscrapers into groups by # of floors, estimate percentages for each. 75% are 30-50 floors (w/ some explanation -> from my experience in cities, most buildings aren't that tall, zoning restrictions, older buildings couldn't have been that big due to technology limitations etc.) 20% are 51-70 floors 5% are 70-110 (tallest buildings are much more expensive, zoning restrictions due to flight paths, only newer buildings could be this tall)
Assign some valuation to each category... Taller buildings more expensive. Seems like a reasonable guess would be to assign value on a per-floor basis... say $50k/ floor.
Calculations 2000 buildings * 75% in 1st category = 1500. Avg 40 floors * 50k/floor = $2M per building. 1500*$2M = $3B in category 1.
Note to interviewer at this point that the building valuation seems like its much too low, since even just houses (albeit nicer ones) can already cost up to $2M. Up the value per floor to $500k and re-calculate.
2000 buildings * 75% in 1st category = 1500. Avg 40 floors * 500k/floor = $20M per building. 1500*$20M = $30B in category 1. Mention to interviewer that this figure ($20M per building) still feels like it might be a little low, but it is much closer to the ballpark than before. 2000 buildings * 20% in 2nd category = 400. Avg 60 floors * 500k/floor = $30M per building. 400 * $30M = $12B in category 2. 2000 buildings * 5% in 3rd category = 100. Avg 90 floors * 50k/floor = $45M per building. 100 * 4.5M = $4.5B in category 3.
Wrap-up Based on my assumptions, total = ~$46.5B real estate in Manhattan.
Mention that this total intuitively still seems a bit low - it feels like the value should be at least $100B but I don't have any strong reason/grounding for where the actual price is, and note some of the areas where might have underestimated in my assumptions... -$500k/floor valuation is most likely still too low, -Manhattan may well be more than 12k acres, -Density of skyscrapers might be higher than assumed (ie. they take up less than 6 acres each), -Distribution of # of skyscraper floors could be much different from my assumptions (ie. there could be more taller buildings), -I didn't take into account that certain buildings would have much higher valuation than normal... for example if its a water-front property, in the center of the financial district, has historical value (ie. buildings like the empire state building or whatever).
Overall, your approach/structure are much more important than your final number. Bonus points for giving some basic explanation / reasoning for your assumptions and for doing sanity checks on your numbers as you go (ie. making sure that your numbers are reasonable and changing assumptions if they aren't).
To be honest, I think your approach is quite poor.
My main criticism is your pricing methodology. Because you have no experience of what the price of a floor should be, you end up with a price range of $50k to $500k, rendering the whole exercise pointless. You should use a price metric more familiar to you, like what the market rate of a 2/3-bedroom flat is.
Secondly, your estimates of area (to find density) aren't especially well grounded. It was good that you used your house as a starting point, but it was suburban, and then you arbitrarily scaled up to find a skyscraper plot size with no real appreciation for the accuracy of that estimate.
Thirdly, you can't say "intuitively" too high/low if your rationale is "it feels like". That's not intuition.
I would have considered a population approach. Start by splitting your analysis into residential and commercial. Mention for greater accuracy you could also include civic. For residential, you can segregate the market into flat sizes, then estimate the price of an average flat of each size, then use the population of Manhattan to find the total market value. For commercial, one approach would be to estimate the number of businesses in Manhattan, segregate by size, estimate the employee density in a building and therefore the space an average business in each size band would need. From here you can estimate the cost of that space, then times up by total numbers of businesses. Do a few checks/adjustments.Done. Should be quicker as well as more accurate.
You think a suburban home covers 1.5 acres? You no concept of what an acre is.
@ Porcine: I went for the most straightforward approach given my (very) limited knowledge of Manhattan. I wouldn't be able to tell you the price of a 2/3 bedroom flat any better than I could price a floor of a skyscraper. Your approach would probably be better for you, but personally I wouldn't be able to even estimate most of the factors you listed in your calculation/approach to either the residential or commercial sections. I'd like to see how it works... any chance you feel like actually typing out your answer in a bit more detail?
Also, what is your definition of "intuition" if not "your gut feeling"? Is that not the context in which I used "feels like"? Just trying to figure out where I miscommunicated.
@TheRapist: I guess you're right... google says average house size = .5 acres. I'd expect an average suburban house to be a little bigger, but 1.5 is probably too big. Oops.
I think the miscommunication on "intuition" was actually mine - I didn't express my point well. What I was trying to say was that you're often encouraged to do sanity checks at the end of each step in your analysis, but you should try and articulate the source of your intuition. It shouldn't be rigorous, but it should be grounded in familiarity. A simple "I know this, so that seems..." would do.
The point about pricing is that 2/3 bedroom flats have a transparent market price, they are traded in those multiples. Whereas a buildings are rarely sold by the floor, so it's a much more arbitrary metric.
I might pad out my approach later.
Wouldn't any of the approaches mentioned above take 10+ minutes of valuable selling-yourself time? In my experience, something shorter is more practical, and leaves room for other questions. I estimated as follows:
Assume Manhattan is 50 x 100 blocks
A block is 1/10 of a mile, so a block is 1/10 * 1600 meters = 160 meters. (interview was for a European office) So Manhattan is (50160) meters * (100 * 160) meters = 5010010101616 = 500,000 * 2^8 = 1,000,000 * 128 = 128,000,000 square meters. Now let's reduce this number by a convenient 25% to exclude streets and sidewalks, yielding 100,000,000 square meters to assess. Note that a square meter is conveniently 10 square feet, so we could jump into feet to make things more practical.
Let us assume Manhattan is 90% buildings with 15 floors, including all skyskrapers, etc.
Big assumption: buildings with 15 floors on average $20,000/sq meter to build.
Is this reasonable? Well it means for small buildings, $500/sq ft, or perhaps $50/sq ft of each floor. Maybe a bit low, but not unreasonably so. This number will be higher for higher buildings, as it should be.
So cost of buildings = 100,000,000 * (0.9 * $5000 + 0.1 * $20000) = 100,000,000 * $6500 = $650,000,000,000
Add in cost of land (100,000,000 * $2,000) and you get $850,000,000,000.
This is probably a bit high, but I don't think by more than a factor of two. I think I overestimated the size of Manhattan a bit, but I was able to do this within 5 minutes while explaining everything to the interviewer.
I can't see how that's quicker than my approach, and again what's the with the obscure pricing methodology?
I think that the process is more important than the answer for this one, so doesn't matter how close/ far you are to the real number. These questions are designed to be answered by anyone not only by those with good sense for real estate prices.
Just structure the analysis (have a reasonable methodology to calculate area and get the math right, show that you understand key dimensions to consider such as the fact that there is unused space such as streets, parks. Show that you know you need to consider different heights and tiers of building for residential, commercial, and other types) and plug in numbers. You are brought in to structure and solve problems not replace databases.
For market sizing questions, are you allowed to use paper? (Originally Posted: 09/14/2015)
Are you generally allowed to use pen and paper to write down the process for how you get to your answer, or do they expect you to figure everything out in your head?
You are allowed to use paper to write down the process before plugging in the numbers. In fact, you should use a paper. Write down your approach first and explain it before you plug the numbers.
You're definitely allowed to use paper yes.
What about a calculator... when you need to divide odd numbers?
No calculator. But can round some numbers. What I did during my interviews is that I ask to the interviewer if it was ok to round, and usually they say of course.
For instance, they ask you to apply 23% tax. Sorry but you will have to compute without rounding. But if it is 23% of 2,455,400. Let's say 2,455,400 is the income before tax. You can round it 2,5 M.
Also you need to learn to compute smartly the numbers. It easier to compute 20% and then 3%. Instead of trying to do everything at the same time.
I hope it helps. You will learn that practicing. ;)
Of course you're allowed to use paper. In fact, when I'm interviewing, I view it as a negative if a candidate doesn't work it out on paper. Sure, some things are easy enough, but given the setting, why wouldn't you use paper? Not worth getting it fast at the risk of getting it wrong.
No calculators though. Usually the math on the interviews is not hard. I have a quantitative PhD, and am very lenient on math mistakes. I'm not going to fail anyone who has a reasonable quantitative background (on his/her CV) for a small mistake.
Interview Question - Finding out the market size for the security market (Originally Posted: 05/29/2011)
im really curious because i got knocked out 1st round.
how would you find out the market size for the security market........... (like CCTV)
i told them each CCTV would need like 3-4 components, and than richer countries have more buildings and money... stuff like, but i couldnt finish lol.
i know the point is the PROCESS... but this is too hard.
thanks!
add up the revenues of all listed security firms for a ballpark figure and multiply by a number less than 10?
i'd say 5 would be a fair amount. That's what i'd do if i only had a few hours, if i had a few days i'd go through the market leaders business reports and accounts and see what they say it is, what it was, get consistent data and average it out.
Ye I'd def come at it from a company valuation angle, rather than a 'count the number of buildings that have cameras' angle. They want to see if you can judge market size by aggregating business activity.
afroman23
what do you mean by aggregating business activity???
thanks
Market Sizing Question (Originally Posted: 03/25/2013)
Do case interview guides have help for market sizing questions? I'm talking about Case in Point, LOMS, or the WSO consulting interview guide
Those do I believe, but you could also simply research these to get the basic idea of how to attack those types of questions. Once you know a few basic techniques its really not that bad.
Case in Point does a really good job. You can also find some examples from Oliver Wyman's Interview Preparation Section under brain teasers.
marketing sizing --> simple = better? (Originally Posted: 09/03/2013)
for market sizing questions, how much depth should you go into with regards to segmenting the question? do interviewers prefer candidates who keep things simple or does more detail = a better mark on the case?
for example, sizing the market for lightbulbs in the US (just households) i just assumed a house has ~10 rooms with 5 bulbs per room. the answer key segmented the house into different room types with the corresponding # of bulbs per room-type. i ended up getting a similar answer at the end, but was i wrong to not go into more detail?
keep in mind this is just one example. I'm asking, in general, is it better to keep things simple?
If you are not doing a bottom-up market sizing and segmenting the market adequately at every level of the analysis, it is rare that the interviewer will be impressed.
When you are doing a market sizing case, there is no right answer, but there are definitely wrong answers. The fact that the answer was similar at the end has no merit.
What you need to do is balance your ability to kill the case with the right amount of complexion when you are segmenting each level of the market, i.e. customers, products, geographies, replacement rates/life cycle, pricing, etc.
It goes a long way doing that one or two extra segmentations, i.e. using only one type of light bulb in your analysis to using three different types of light bulbs, how each different type of light bulb differs in price, replacement rate, etc. - this one example, there are many others.
Simple is not better, but if you are going to complicate things make sure you can still do the math flawlessly. If you can pull that off, you're in good shape.
50 lightbulbs in the average US house?
Hard to give a specific answer, but not too simple that it takes a few seconds to answer and makes it look like you aren't trying or don't get it, but not too complex that will take over 3-5 minutes to explain or will be hard to get all the math right.
I think the emphasis should be to consider the key variables that would impact a real world case: replacement rate, price tiers/ types of bulbs, mix of house/apt sizes. There is no right answer and sometimes you can choose 2-3 dimensions and do the math on those and then say that if you were on a real case, i.e., had time and data you would look into some additional ones. What they would test there is that you indeed prioritized the key ones and could still think of others.
Perhaps a good way to go about it is to take one example, like your house one and really flesh it all out. Play around with it a bit and what level of detail really moves the needle will become more apparent. This is just like real life: as you practice you learn to distinguish what's truly important and do so in an increasingly efficient manner
Market sizing/logic questions in lower tier firms? (Originally Posted: 10/04/2010)
Think IT or operations consulting at a firm which recruits MBAs primarily outside the top 10 schools. Do I have to worry about preparing for stuff like market sizing?
Quae quis nobis fuga blanditiis quibusdam. Quod provident officiis repellendus voluptatibus in dignissimos non.
Quis laboriosam eveniet cumque nesciunt quis harum voluptas itaque. Quis vel et ea a quidem dolorum voluptates sunt. Quia pariatur eum maxime et. Voluptas quos atque recusandae numquam quis debitis.
Veritatis earum cumque quis enim impedit ducimus. Et est unde cumque quam impedit ipsa alias nobis. Perferendis doloremque omnis qui fugit voluptatem. Voluptas et cum eius ab maiores magni. Provident eos totam laudantium possimus consequatur quia harum vitae. Eos iste dignissimos est impedit doloremque.
Necessitatibus ipsam rerum numquam molestiae laborum. Aliquid dolore molestiae est. Voluptatum est expedita vel. Aut voluptas non nemo sequi accusantium aut velit.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Et et harum voluptates delectus unde aut aut. Eum consequatur sapiente et repudiandae autem. Totam officiis nobis ipsum id. Nisi magni nesciunt aut hic nesciunt rem.
Modi animi sequi sequi nulla sunt. Eos pariatur pariatur laboriosam ipsa quisquam. Dolorum omnis soluta aut deleniti assumenda mollitia deserunt voluptates. Impedit eos doloremque qui autem soluta quia. Vel impedit iure voluptas excepturi officiis ab fugiat.