TAL Education Group (TAL) 2018 Q3 法說會逐字稿

Ladies and gentlemen, thank you for standing by, and welcome to the TAL Education Group Third Fiscal Quarter 2018 Earnings Conference Call (Operator Instructions).

I must advise you that this conference is being recorded today, Thursday, 25th of January 2018.

I would now like to hand the conference over to your first speaker today, Ms. Mei Li.

Thank you, and please go ahead.

Thank you all for joining us today for TAL Education Group's Third Fiscal Quarter 2018 Earnings Conference Call.

The third fiscal quarter earnings release was distributed earlier today, and you may find a copy on the company IR website or through the newswires.

During this call, you will hear from Chief Financial Officer, Mr. Rong Luo.

Following his prepared remarks, Mr. Luo will be available to answer your questions.

Before we continue, please note that the discussions today will contain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995.

Forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from our current expectations.

Potential risks and uncertainties include, but are not limited to, those outlined in the public filings with the SEC.

For more information about these risks and uncertainties, please refer to our filings with SEC.

Also, our earnings release in this call includes discussions of certain non-GAAP financial measures.

Please refer to our earnings release, which contains a reconciliation of the non-GAAP measures to the most directly comparable GAAP measures.

I would now like to turn the call over to Mr. Rong Luo.

Thank you, Mei, and good evening and good morning to you all, and thank you for joining us on this earning call.

I'm pleased that we managed to maintain our strong growth momentum in the third fiscal quarter 2018.

Our top line growth continue to be driven by high demand in all cities and the fast ramp-up of our online courses.

Top line growth in the quarter was 66.3% year-over-year in dollar terms to USD 433.3 million.

In renminbi terms, net revenue grew by 62.7% year-over-year.

Enrollments increased by 85% year-over-year.

GAAP income from operations increased by 102.1% to USD 44.6 million in the third quarter, and a non-GAAP income from operations increased by 82.6% to USD 56.8 million.

As expected, during the third quarter, we are gaining leverage from both the new classrooms as well as new teachers added in the previous quarters.

As they gradually [come] to full use, we are happy to see the pressure on the year-over-year comparison of gross margin reduced compared to that of the second quarter of the fiscal year.

Non-GAAP operating margin was up by 120 basis points year-over-year in the quarter.

Now Mei will give an update on our operational progress in the third quarter.

After that, I will update you on our ongoing investment in technology, which is guided by our long-term strategy vison of advancing education through science and technology.

The robust revenue growth in the third quarter was from all business lines in all cities.

Small class, which consists of Xueersi Peiyou small class, FirstLeap, Mobby and some educational programs and services, accounted for 82.8% of total net revenue compared to 84.7% in the third quarter last year.

Peiyou small class, which remains our core business, represented 72.9% of the revenue compared to 74% -- 74.9% in the same year-ago period.

This lower revenue contribution from the small class was mainly due to the faster growth of online course business, which accounted for 7.8% in the quarter.

Net revenue from Peiyou small class was up by 62.4%, while enrollments increased by 91.7%.

In order to have our existing Peiyou students to learn based on varied individual aptitude in each city, we encourage them to take additional Xueersi Peiyou online course that allows them to easily take more subjects and reinforce the learning processing of flexible manner.

Peiyou off-line small class continue to run healthily.

At the same time, we are very happy to see the significant enrollment contribution from Peiyou online after the summer term.

We will continue to roll this out into more cities.

Xueersi Peiyou small class revenue generated from cities other than the top 5: Beijing, Shanghai, Guangzhou, Shenzhen, Nanjing, grew by approximately 60%.

This is other than the top 5 accounted for 39% of the Peiyou small class business, almost unchanged from the same quarter last year.

Revenue from top 5 cities grew by 62% year-over-year, faster than the other cities, mainly because of the contribution from Peiyou online course.

We have achieved a triple-digit year-over-year renminbi revenue growth in 9 out of 27 cities that we entered by the third quarter of fiscal year 2017, including Shijiazhuang, Changsha, Qingdao, Luoyang, Nanchang, Ningbo, Hefei, Wuxi and Xuzhou.

The revenue growth across all cities was driven by incremental ramp-up of enrollments from our earlier classroom expansion.

Enrollments from Chinese classes and English classes grew faster than the other subjects as a result of our continuous efforts to roll out these courses.

In the fall term, they accounted for over 20% of enrollments from Peiyou small class.

And as of the winter term, we started to teach Chinese classes in 13, that is 1-3, cities and English classes in 20 cities.

We will continue to expand subjects in more cities and make our business more diversified.

We'd like to remind you of the impact of Chinese New Year on our course scheduling in the fourth quarter.

Last year, we had approximately 2 classes of [running up to] spring term course booked in the fourth quarter of the fiscal 2017 because of the early timing of Chinese New Year at that time.

In contrast, our next spring term course will start from March 2018 in the first fiscal quarter because the Chinese New Year is coming late this year.

Since we recognized revenue proportionately after we deliver each class, no revenue from spring term course will be booked in the fourth quarter.

At the same time, the negative impact from the late start of spring term will be somewhat offset by the smaller portion of revenue shifted from Q3 to Q4 due to a late start of the recent fall term.

Again, this is the cutoff issue arising from seasonality and is expected to benefit the revenue growth of the second quarter of fiscal 2019, in which the spring term will end.

Later, we will give the revenue guidance for fourth quarter.

For one-on-one business, we had a healthy third quarter and achieved a year-over-year revenue growth of 51.4% in U.S. dollars.

One-on-one, including the overseas consulting business, accounted for 8.1% of total revenue compared to 8.9% in the third quarter of fiscal 2017.

Turning to our capacity expansion.

In the third quarter, we slowed it down as planned and added 16 new learning centers and closed down 12 learning centers, adding a net 4 learning centers.

During the quarter, we added 68 Peiyou small class classrooms on top of the 2,021 we had added in the first half of the fiscal year, making the number of small class classrooms increase by 47% compared to the number by end of November 2016.

We added most of the small class classrooms in Beijing, Guangzhou, Nanjing, Xuzhou, Luoyang, Nanchang, Ningbo and Hefei.

Following the routine room renovation and the necessary air cleaning, those classrooms will gradually come into use in the later regional utilization, approximately by the summer term.

Meanwhile, we continue to expand into new cities at pace.

In the third quarter, we entered 2 new cities, Shaoxing and Yangzhou.

During the quarter, we added a net of 5 small class learning centers and a 2 FirstLeap small class learning centers.

We closed a net of a 3 one-on-one learning centers as part of our standard performance review.

By the end of November, we had 579 learning centers in 38 cities across China, of which 414 were Peiyou small class, 8 were Mobby small class and 60 were FirstLeap small class and then 97 were Zhikang one-on-one.

Looking to Q4, we will maintain well pace across the expansion, with the estimated addition of 10 to 15 Peiyou small class learning centers.

Moving now to our online business.

Third quarter revenue from xueersi.com grew by 175% year-over-year.

The growth momentum of online continues to be very strong due to the success of live broadcasting.

Online contributed 7.8% of total revenues this quarter compared to 4.7% in the same year-ago period due to the fast growth of online live course.

As a first mover in going live online on xueersi.com, we are highly confident about the long-term opportunities to live broadcasting.

Finally, on the revenue side per business line, other revenue is mostly from the online advertising business and represents 1.4% of the revenue in third quarter compared to 1.7% in the third quarter of fiscal year 2017.

Let me now go through some key financial points for the third quarter of fiscal year 2018.

In the third fiscal quarter, small class ASP decreased by 15.3% in U.S. dollars year-over-year.

The decrease was mainly caused by significant enrollment contribution from Peiyou online, of which the price is approximately 40%, that is 4-0 percent, lower than Peiyou off-line classes.

Meanwhile, ASP of Peiyou normal offline small class increased by high single-digit percentage in the third quarter.

Zhikang one-on-one ASP in U.S. dollars increased by 12.1%, because of the normal price increases mostly, started from the fourth quarter last fiscal year.

Online course ASP increased by 61.7% in U.S. dollars in the third quarter, driven by the fast expansion of the online live class.

Cost of revenues increased by 67.7% to USD 221.1 million from USD 131.9 million in the same quarter 1 year ago.

The increase in cost of revenues was mainly due to an increase in teacher compensation and rental costs.

Non-GAAP cost of revenues, which excluded share-based compensation expenses, increased by 67.6% to USD 221.0 million from USD 131.8 million in the same year-ago period.

In the third fiscal quarter, gross profit was USD 212.2 million from USD 128.7 million in the same year-ago period.

Gross margin for the third quarter was 49% as compared to 49.4% for the same period of last year.

Operating income increased by 102.1% to USD 44.6 million.

Non-GAAP operating income increased by 82.6% year-over-year to USD 56.8 million.

Other income was USD 5.4 million for the third quarter of fiscal year 2018 compared to other expenses of USD 0.7 million in the third quarter of fiscal year 2017.

Other income in the third quarter of fiscal year 2018 was mainly due to gains from the disposal of investments.

Impairments loss on long-term investment was USD 1.5 million in the third quarter of fiscal year 2018 compared to USD 3.8 million in the third quarter of fiscal year 2017.

Impairment loss on long-term investments was mainly due to the other-than-temporary decline in the value of long-term investments in several investees.

Income tax expense was USD 11.4 million in the third quarter of fiscal year 2018 compared to USD 3.1 million in the third quarter of fiscal year 2017.

The increase was mainly due to increase in income before provision for income tax and loss from equity method investments.

Basic and diluted net income per ADS were USD 0.08 and USD 0.07, respectively, in the third quarter of fiscal year 2018.

Non-GAAP basic and non-GAAP diluted net income per ADS were USD 0.10 and USD 0.09, respectively.

From the balance sheet, as of November 30, 2017, we had USD 1,126.6 million of cash and cash equivalents and USD 184.3 million of short-term investments, compared to USD 470.2 million of cash and cash equivalents and $229.5 million of short-term investments as of February 28, 2017.

Capital expenditures for third fiscal quarter were USD 28.3 million, representing an increase of USD 11.2 million from the USD 17.1 million in the same year-ago period.

The increase was mainly due to leasehold improvements and the purchase of servers, computers, software systems and other hardware for the company's teaching facilities and mobile network research and development.

As for November 30, 2017, the company's deferred revenue balance was USD 1,074.9 million compared to $679.9 million as of November 30, 2016, representing an increase of 58.1%.

Deferred revenue primarily consisted of the tuition classes we advanced for the fall semester, winter semester and the spring semester of small class, as well as deferred revenue related to other businesses.

Now I will hand the call back to Mr. Luo to highlight our recent progress in exploring science and technology and the business outlook of the next quarter.

Thank you, Mei.

Now I would like to give you an update on our progress on the technological investment.

As our long-standing record shows, TAL has been deeply committed to exploring the possibilities of technology-based education reform.

We will continue to ramp pilots in AI and other cutting-edge areas of EDTECH, to create innovative ways to better serve our customers.

As part of our continuing investments in the new technology, we have upgraded on our Xueersi Intelligent Practice System, IPS, to Xueersi cloud-based learning platform that was launched last week.

The new platform has upgraded the functions, contents and system architectures of the current IPS.

We've made the learning personalized and customized and gives our students access to both online and off-line experiences, creating a complete [closed loop] of learnings.

We recently announced a long-term equity investment firm agreed to purchase a total of USD 500 million of the newly issued Class A common shares of TAL.

The transaction has been completed in January, and we will use the net proceeds of the offering for general corporate purpose, including funding our plans to create the innovative applications of the smart technologies in education context and to build the revolutionary new models of learnings.

Following these transactions, the investor holds, taking into account its existing holding, approximately 5% of the company's outstanding shares.

Let me now move to an outlook for the next quarter.

Due to the timing of Chinese New Year in the fourth quarter fiscal year 2017, we recognized approximately around 10% of spring term revenue, but 0 in current fourth quarter.

After netting out the 2% fall term revenue shifted to the fourth quarter due to the late start of the fall term this fiscal year, based on our current estimates, total net revenues for the fourth quarter of fiscal year 2018 is expected to be between USD 474.5 million and USD 480.8 million, representing an increase of 50% to 52% on a year-over-year basis.

This cutoff issue arising from [this analogy] will impact both top line and margins.

These estimates reflect our current expectations, which is subject to change.

That concludes my prepared remarks.

Operator, we are now ready to take questions.

(Operator Instructions) Your first question comes from the line of Alex Liu of Daiwa Securities.

First, how many of the classrooms do we add this quarter?

And for the small class enrollment, does it include both online and off-line, right?

And what's the growth look like for the off-line courses for the small class segment?

And my second question is on capacity.

I think given our capacity growth continues to decelerate this quarter, should we expect some kind of a cyclical acceleration for capacity in the coming year?

Or simply, the capacity will be more or less stabilized at certain levels in the coming 1 year?

Thank you, Alex.

Your first question about how many classrooms were added for the Peiyou small classes, actually, this quarter is around 8 -- 68, 6-8, classrooms compared to the 2,021 we have added in the first half of fiscal years.

I think as what I have mentioned in the previous call, actually starting from Q2, we start to hold a little bit in the capacity expansion, and we need some time to digest the capacity we have added the previous [4] quarters, so this will continue in Q3.

And looking forward to Q4, actually we started to accelerate a little bit in Q4.

Our current outlook is we'll -- probably we'll add around 10 to 15 new learning centers in the Peiyou small class business.

So for the full year, maybe we're adding around more than 30% in capacity [perspective].

And because in the past few quarters, we made some progress to include -- increase the seat fulfillment rates by -- in our classrooms quarter-over-quarter, but considering we have added a lot in the past.

So, in the year-over-year conversion [of studies], we -- even today, we are still below.

So we have the [going up] potential even we only add a little bit in Q3, but which we'll still continue to give us very strong space for us to deliver the healthy growth next year.

And you're also asking about online and off-line revenue growth, so let me clarify this actually.

And on the online, means the Xueersi online school, which is growing quite well last quarter, it grows around 144%.

This quarter, they grew around 175%, and we foresee they will continue with grow -- very strong growth momentums in the coming year.

Off-line is also very stable.

I think their net revenue from the Peiyou small class segment, that was up by 52.4% this quarter.

And looking forward, we are quite confident off-line business will continue to be very healthy and very sustainable and very stable.

And on the capacity for the next year, I think as we have mentioned in the past, we as a company, we definitely -- we would like to manage the healthy growth.

So we think an ideal range for us will be around 30% to 50%.

And last year, we're adding along, I think around 80%.

This year, we controlled a little bit in the pace, around 30%.

So looking forward to the future, we will go back to the right range of 30% to 50% in capacity perspective . Besides the more classroom numbers we have added, actually, we need to pay more attention to how we could leverage these spaces to give the quality of services to the students, how to increase the healthy KPIs of our -- or represents, to sure make we are running everything on track.

So we care more about quality and we care less about the speed of expansions now.

So looking forward, we'll go back to managing healthy growth.

Thank you, Alex.

Your next question comes from the line of Sheng Zhong of Morgan Stanley.

(foreign language) A little bit follow-up about capacity expansion.

So can you give more color on the capacity on how much percent will be like dual-teacher model among them?

And also, by the way, can you give more updates on the -- how is the dual-teacher model (inaudible), like what the rotation rate and what's your expectation for the dual-teacher into next year?

And second question is about the online business.

Can we give more guidance on the online business revenue and enrollment outlook for next year?

And also do we -- how -- just want to understand how -- for the efforts will the company make to on the online business, including the technology investment?

And also, do we have any significant promotions on these online business?

Thank you, Zhong Sheng.

Your first question about the capacity for the dual-teacher models, what I can say is actually the new cities we have entered, most of them are dual-teacher model cities.

And we continue to add more dual-teacher classrooms in current cities and in new cities.

And this year, I think the contribution from the dual-teacher model is still minimal, below 5%.

And we are confident next year, especially I think maybe around next year, summer, we're probably going to see a much better number at that time.

And for the online, I think our vision in online is that it deliver very fantastic results.

I think the whole year, the revenue growth was more than 140%, and we see this quarter is 175%.

And looking forward to next year, we foresee at least they can keep their growth momentum to be more than 100% in revenue growth perspective.

But when we talk about the online, we need to keep in mind, the most important play for online actually is we need to get more market share.

Online is a unique business, very different from the Xueersi Peiyou off-line business.

The online, they have different targets and they have different structures.

And we -- when we want to get more market share, we have 3 things we need to be very careful of.

The first one is we need to continue to invest in technology.

So on one side, we could to hire more people, more teacher assistants and more salespeople to sell our online products.

But because we strongly believe the technologies are the only game-changers in education market, so the same amount, if we have to invest on hiring more people and more staff, we'd prefer to spend them in technology and platform so we can try to reduce the workload of every single people.

We can maximize the peoples' efficiencies, which can focus to improve the overall profitabilities and efficiencies of the whole business models.

The second thing is we need to invest actually is marketing, both online and off-line.

I think maybe some of you have -- saw our off-line advertisements in some selected cities in the past, I think it was January.

And the reason is because Xueersi online school actually, in nature, they have the much bigger and wider market potentials, and we have a lot of white space markets we've never touched before.

Even in the big cities, we have a lot of sub areas.

And some of the places, we don't have a presence at all and we have huge potentials over there.

So -- but Xueersi online school also need some right way to convey our messages to the customers.

So we have tried some online and off-line marketings, especially in online strategy.

But of course, we will balance the return of investments in the marketing strategy to make sure we don't lose money.

And the third thing we need to be very careful is about the operating models.

When the online business is growing so fast, 175%, the team size is also increasing.

Even we have seen the revenue per head count is increasing, we still need to be very careful how we can design the growth structures to manage a much bigger size in the people perspective.

So the operational model, how we can upgrade or maybe evolve that new operational model to cover more, that is something we need to be very careful.

Even we are the first company, maybe one of the first company in this industry, which we try the online school I think 5 years ago, but we still need to be very careful.

There is nothing easy.

So looking forward in general, we are still quite confident about that, but we will also be very careful on the executions.

Thanks, Zhong Sheng.

The next question comes from the line of Natalie Wu with CICC.

Two questions.

First on Peiyou online courses.

Just wondering how many enrollments done -- of Peiyou online courses done are coming from off-line Peiyou small class students this quarter?

And how many cities does TAL Peiyou online courses cover currently and what's the target of that this calendar year?

And also, what's the teacher's revenue sharing percentage for that business?

Just want to get the rough [billing] of the margin prospect as -- in longer term.

Second one is on the R&D expenses.

I noticed that you have set up the TAL [bring] lab today.

And also you mentioned that you will grow your R&D team from current 4,000 to 10,000 in next 3 years, including 500 AI team.

So just wondering can you still control your original R&D budget as to the 5% of revenue scope for that?

Thank you, Natalie.

I think the first one, the question about the Peiyou online, actually which is part of our pilot we were running before, we called it [hybrid] at the time.

So we combined hybrid to Xueersi Peiyou to -- then changed the name to be Peiyou online.

So actually, you know the Peiyou online is quite different from Xueersi online school.

Xueersi online is quite unique and quite independent from the Xueersi Peiyou, while the Peiyou online will highly connected to their Peiyou off-line classes.

I show one example is -- for example, we went to our classrooms in Beijing.

In the classrooms, the teachers use the iPads to -- or IPS to collect a lot of data from students.

And when we analyzed this, all the data, we maybe figured out some important knowledge points.

We have certain percentage of students who have challenges, so we quickly opened the kind of the shortened class of the -- through online to teach them.

So this become a very efficient way to give complementary services to the students who have challenges or who have something they need to overcome from the off-line [teachers].

So in -- this become a very good way for us to increase the values per customer.

But in nature, we still wish to leverage these offers to serve the students much better.

Specifically, in margin perspective, because we don't have any customer acquisition cost for Peiyou online and the online is quite similar to the Xueersi online school, actually we don't have that many teachers for that.

So when everything is on track, that will be kind of beneficial to our total margins.

But again, Xueersi online will only try that for 2 quarters.

The first quarter is in summer, in the fall is the second quarter.

So we probably will need more time and more time to see what it will be in the future.

We personally believe the online will be a very good add-on of our services to the students.

It can create a white space market for us.

Let's stay tuned for maybe the next quarter and next year to see what the final result will be.

It's positive, but we still need a bit more times.

The second question about the R&D side.

Yes, we are very happy to see we have set (inaudible) [brand sense lab] today.

And when we talk about R&D actually, we need to go back to the time 5 or 6 years ago when we're one of the first companies in the industry to try their online school.

Their -- at that time, that was prerecorded content.

And 2 and to 3 years ago, we also only invest in technologies to transform the prerecording online school to the live online school.

And at the same time, we tried their other hybrid models.

And we also leveraged all the learnings and the knowledge we get from online and we start to establish the dual-teacher models maybe 2 years ago.

So we continually investment more and more over here.

And last year, we tried some new technology in AI perspective, including the facial recognition technologies and the voice technologies.

And this will become a continuous effort in our companies because our mission is try to advance education through science and technology.

And about how much money we spent on this?

Actually, we have a very good discipline.

Every year, we spend the similar percentage of our revenue on the R&D side.

And for some special projects, we'll have some special funding.

But in general, the percentage of revenue spending and R&D are quite similar.

Thank you, Natalie.

Your next question comes from the line of Mariana Kou from CLSA.

My first question is actually on the margins.

And how should we think about longer term?

I know we're talking about capacity of about 30% to 50% growth in -- on an annual basis longer term.

Should we expect some sort of margin operating leverage still?

Or how should we think about that given that, now, online is quite strong for you?

And then you just kind of talk about the Peiyou online as well, that seems to be quite good for margins given that the customer acquisition cost is quite minimal.

And my second question is just broadly on technology.

I think we talked quite a lot about the technology investments and how we're trying to incorporate more technology into the learning experience.

Could you share a bit more color on management's view on whether we will, going forward, think about more M&A, or more self-development or just sort of partnership, and especially in light of the recent kind of long-term investment that we got, the $500 million?

Thank you, Mariana.

The margin question -- okay, thank you for the margin question.

I suppose that should be the first question I should answer.

And last quarter, we have guiding The Street is we have some challenges in margins, especially in the second half.

And the whole year definitely, we also will slightly decline in the margin by around maybe low single digits last quarter.

And coming to this year, we see some positive news coming from our Q3 performance, on one side, we start to see the capacities we have added in the previous quarters now gradually coming into use.

Their capacity utilization rate is improving quarter-over-quarter.

Even [compared] to last year, they are still below, but quarter-over-quarter is improving.

The second reason is because we also see our newly added -- our newly acquired business actually the Q2 is the big season for them, which I have mentioned last quarter so many times.

And so relative to big season in Q3 and Q4 compared to Q2, they will be much better than that.

And we also see some leverage coming from the one-on-one business and the Xueersi online school business.

So in general, finally we've got a very good result, better than what we expected in the last quarter earning call.

But looking forward to Q4, we also have to mention, because of the seasonalities we talked about the 3 -- 2, 3 times in the prepared remarks already, this will shift the -- around 2 weeks of our revenues from Q4 to next year Q2, which is a lot of money.

So this shift will impact my -- both top line and bottom line.

So in Q4, our bottom line, our margins will decline.

And for the whole year, we continue our guidance to The Street last quarter is around -- our OP margin will be declining by around 1% to 2%.

So we don't change our views, that's still a range we will be for this year.

And looking forward to next year, and the situation has changed a lot, in the first place, let me walk through the segments one by one.

For the Xueersi Peiyou business next year, definitely, we need to manage them back to healthy pace.

So we will require the Xueersi Peiyou business to improving in the key KPIs to deliver very healthy growth.

But the Xueersi online school, actually, there's a market share gain.

If we spend more time to deep dive into the population or the demographics of China, actually we can see there is a huge opportunities for the younger-aged students that are moving from off-line to online.

Now they are more and more -- getting more and more familiar to the online solutions now, which is also proof of our online datas.

So in this study of next year for online, that is a market share gain.

So we need to invest in technology, in marketing channel, in operating model and all of that to try to get more market share in online perspective.

And so, in general, I think next year margin, I probably I can give you much clearer colors in next quarter earning call when we finish our Q4.

But in directional thinking -- directional perspective, I think we are still feel confident about next year we could maintain our margins at least flat.

And once there -- and once the final numbers will be delivered, probably I'll give you more colors and more information on what visibilities in Q4 earning call.

And your question about the technologies that are available through M&A or we invest more on our in-house development, what I can say is actually the technology sometimes is very difficult to buy for anyone.

So in general, when we will want to [aim] -- for education, sometimes is a very vertical market.

So the big players, they even don't have much time to do a specific R&D on education.

So we have to rely on ourselves a lot.

So that's part of the reason why we built a very big team on IT and product strategies, and we will continue to add more people to the team to develop some new technology and new system.

For example, like the dual-teacher models and Xueersi online school, all the solutions actually is the best.

But at the same time, if some of the technologies, which if they're cutting-edge technologies and the team's very good, we are not -- we are not reluctant to buy them.

And we are very happy to see some of the good teams, they have very good potential and they are willing be part of us, we are more than welcome they -- to let them join our big family.

But in general, we need to focus our in-house development first as well as looking into new possibilities in the whole industry.

Thank you, Mariana.

Your next question comes from the line of Ivy Luo of Macquarie.

My first question is a follow-up on our online segment.

So just wanted to understand, just we know the margin improvement this quarter is from both utilization and our online business that's growing really fast.

But what's the margin like for our online segment?

And we also mentioned that we're going to do some marketing and promotion to gain market share in online space.

Do we have a budget in terms of the marketing dollar or R&D that we're going to spend in turn to grow our online business?

And also I want to understand what's the -- how much of the online business is currently from our, say, top 5 cities?

That's my first question.

Thank you, Ivy.

Let me take some notes first.

Okay, your first question about online margins.

The Xueersi online school, they are profitable this year, and what I can say is it's a meaningful, profitable this year.

But looking into the next year, we have to say that online is our market share gain.

So in the directional perspective, we are willing to spend more money on online perspective, in technology, in marketing, in new operating models.

So we don't want to increase the margin for online next year, we are willing to pay more money for online to try their best to win more market shares.

Of course, we will not let them to lose that much money.

But compared to the other segments, we will give more possibilities to online.

While off-line, there's Xueersi Peiyou small classes.

This year, we suffered a little bit, I think, because we have added a lot of classrooms in the year beginning.

But we fully believe next year, the Xueersi off-line will go back to normal, we will steer back to our healthy on track.

So in that perspective, we could see some possibility to get some margin improvements from that mainstream business.

So overall, as a company, we are still -- we are quite confident about margin for next year.

And about your question about how much of my online customer is coming from top 5 cities.

What I can say is because online, actually, they don't have any barriers in geography perspective.

So online today cover a much bigger geography coverage than off-line.

And they're -- we don't see the very high numbers, as high as the Xueersi off-line schools, to show up in the online school revenue.

And looking forward to next year, actually we wish that online can not only penetrate the Tier 1, Tier 2 cities but also have some possibilities to penetrate more space step by step.

So we don't hurry to harvest the market overnight.

We need some patience.

We need to let the team to continue to tweak their models to mature and find the right channels to penetrate more students and continue to improving the teaching quality, continue to improving the efficiencies of operating models.

For example, let one -- let every teacher assistant can cover more students at the same time.

So we still have a lot of things that we can do.

And we care less about how many people coming from certain cities in our online business, we -- what we want to cover is the mass market.

So we probably can give you more colors maybe next quarter to see online perspective because we've piloted some marketing channels, both online and off-line, in January, so we need some time to analyze the results.

Thank you, Ivy.

What is your second question?

My second question is on your promotion.

I think you mentioned that we're going to do some winter promotions now.

So can you give me some color on the -- like the winter promotion?

And also a housekeeping question.

I understand our online revenue grew 170% this quarter, how about the online enrollment number?

Sorry if I didn't hear that number previously.

Okay.

This is looks [very] [rough], but I think the online enrollment is around 21% of my total enrollment, and the online enrollment growth is around 74%, 75%.

And about promotions actually, for the off-line business, we don't have that many promotions, and we benchmarked some of our counterparts' performance and their plans.

We probably can see next year, there -- possibly there will be a lot of promotions from our counterparts.

But we as a company, again, we still believe that teaching quality is the only criteria for the parents when they make decisions.

And the teaching quality is the only guarantee to the students when they pay and come to our school.

They invest their time and money on us.

So we wish that's the only good things we can show them as a return.

So looking forward to both in the winter and in next year, we don't have any mass plan on promotion perspective.

We will continue to be very cautious about the promotions.

And for the online school, actually, online school promotion will be different issues because online is a totally different experience for the students.

So sometimes, we're running some pilot classes.

For example, around maybe 1 or 2 classes, we'll only charge them around RMB 50 in online.

We persuade them to try the online first and then we convert them to our regular classes.

So for online, when we want to get -- when we want to get more market share, we will continue this approach.

But for off-line, we will be very cautious.

We don't have any plans to make much a bigger promotion next year, and we will be very cautious and we'll continue to improve the teaching quality as the first priority.

Your next question comes from the line of Fan Liu of Goldman Sachs.

So I have 2 quick questions.

So the first question is about your acquired business, including the Firstleap and Shunshun.

Would you mind sharing with us more color on maybe the operating metrics or the financials numbers, profitability for these 2?

And also the second question is about our R&D this quarter.

So basically, I think the operating leverage mainly come from this line.

Would you mind to share with us any color on that?

Thank you, Liu Fan.

Firstleap and Shunshun are in different situations.

For Firstleap, actually they're running quite well and the revenue growth is quite healthy, as well as their bottom line margins.

So for the Shunshun, actually we have mentioned last quarter, because they need to pay for some of the costs they have, they have done last year, so Q2 is a bit lost to them.

Q3 is still loss-making position, but compared to Q2 it's much better.

So currently, we're running some of the optimizations in the Shunshun.

But again, the key purpose for all of this is, in the first place, we need to focus on our customer first.

If in some places, we can now provide the best quality services, so we should now hurry to increase the size of the business over there.

So we need to always put their customers as first priority.

Secondly, I think is similar as their -- our Xueersi [study], we prefer to be being stronger than to be being bigger.

So we want to focus a little bit on the key cities and on the very important business lines.

For example, the major cities; and the very important business lines like the U.S. lines, the U.K. lines, to make sure we can deliver a very robust and strong business models.

Before we want to -- we go to any bigger.

So we fully believe the team from Shunshun and the team from FirstLeap will continue to perform quite well.

And next year, we'll probably going to see some surprises, good surprises coming from them.

Secondly, about R&D expenses.

Currently, we don't have any unusual R&D investments in the past quarters.

And we will maintain the similar percentage of revenue to invest in R&D.

And this year, we focus a lot on AI and brand sense, which may now deliver some real results, while shortened, but very important and very critical for us to establish the new education models in 3 years' time.

If we can recap what happened 3 years ago, when just investing online, when we're just investing in the live broadcasting platform and we were just investing in the dual-teacher models, some people may say that's unnecessary.

The final result is if we don't do any investment at that time, coming to today, I'm still an off-line company.

So we will balance the efficiency and the profitability of the investments of R&D and we will continue our focus in this area because our mission is to advance education through science and technologies.

Thank you, Liu Fan.

Your next question comes from the line of Thomas Chong of Credit Suisse.

I have 2 quick questions.

The first one is about regulatory updates.

Should we see more regulations in coming years?

And my second question is about the guidance.

Can you just briefly talk about the ForEx impact and most importantly in RMB terms?

Okay, Thomas.

The first question about regulations.

I think among all education companies, we are the only companies who have talked about that for 5 quarters about already.

And so again, in the first place, we're working all the regulations and policies released from the governments, and we're 100% follow the government requirements to continually improve our operations.

Last year, Congress has approved the new law.

And today, this year, that's the time for different province, they need to carry out the new laws into the [right detailed] policies.

We'll start to see a lot of province like Beijing, Shanghai, Tianjin, in a lot of places, they have come out their results.

And all of these policies -- actually, the directions is very clear, is we need to make sure we have some -- they try to increase their policy requirements on the tutoring companies because today, the market of tutoring companies actually is very fragmented.

Some of them even worries more.

So sometimes, they will lead to some potential risk, maybe in safety risk, maybe in the other kind of risks.

So I think with the new policies coming out to try to regulate this market and raise the bar for this market, we as the leading player, we will definitely welcome all regulations and try our best to fit into the new models.

And specific for [Xueersi] perspective, actually we deal with issues from 5 quarters ago, so we have set out our specific organizations, specific team, our own risk efforts.

And today, our teams is working very closely with the different province and city level governments to carry out all their new policies, which have been launched maybe in the past few months.

And we don't see that is -- we personally see that is a good opportunity for us to increase our operations.

And maybe, potentially, we help to increase the consolidation speed of these markets for the top players.

And your second question about our guidance on Q4, specifically in RMB.

I think our guidance on the dollar terms is 50% to 52%.

This year, we consider around 7% to 8% in their currency depreciation perspective.

And looking forward, I think next year, their potential should be -- potentially, there should be some kind of currency depreciations over there.

But I think, as a company, sometimes we don't care about the currencies.

We need to be very straightforward with ourselves, that's what is our real growth.

And looking into the next, well, 5 years, we are still quite confident that the overall company will deliver top line revenue growth in the range of 30% to 50% in -- no matter what currency perspective.

So we will continue to carry out our way to manage a healthy growth.

And yes, that's it.

Thank you so much for your question.

Ladies and gentlemen, this concludes today's conference call.

Thank you for your participation.

You may all disconnect.