NTES – 1Q 2009 revenue might be weaker than predicted

I had been tracking the server statistics on NTES’s game for a while now. I tried to use that statistics to predict the quarterly earnings for NTES. But my last prediction was horrible. I was off by $11M in my prediction for the last quarter, please see the following article:
http://chinese-net-gaming-stock.blogspot.com/2009/03/ntes-4q2008-revenue-prediction-error.html

The reasons for the bad predictions are two folds. Firstly, I don’t know how to take account of the seasonal factor between quarters. I use last quarter’s ACU (Average Concurrent Users) number as a base to compare to this quarter’s server counts. That provides me with an estimate of this quarter’s ACU. But there is a seasonal factor between quarters. At this point, I am using trial and error to come out with a seasonal factor that can accurately capture this effect.

The second reason for my bad prediction last time is that I was confused about closed beta status of NTES’s two new games, TX2 and XYW. Both had achieved significant success. Both games gets about 100k to 150k players playing consistently and are still growing. But both games are only in closed beta. NTES already monetized them. Players can buy items within the game’s virtual store.

For every other companies, they would have commercialized these two games long time ago. But not NTES, she is so deliberate. Items within the virtual store has very little power and therefore are only there as part of closed beta test. It is there to test the working of the virtual store and the items themselves, but not to make money.

For a 100k-150k item based game, one can easily get $8M per quarter. I thought NTES started commercialize these two games. When NTES said that these two games are still in closed beta, I thought it can’t be right because other companies would had commercialize the game (that already achieved 100k to 150k players).

Because of the above two factors, I would not provide quarterly revenue estimates.

But according to my server count, I observed a lower server count in 1Q2009 compare to that of 4Q2008. According to analysts estimate (in Yahoo), 1Q2009’s revenue is $116M. That would only be $1.5M lower than that of 4Q2008.

Somehow, I think 1Q2009 revenue will be hard to get to that level.

Financial crisis clearly had a affect on China’s gaming sector. It is a very mild affect, but still a negative effect. It is also a temporary effect because according to my server count, the games are starting to grow again in April.


But the wild cards are TX2 and XYW. Both games are popular and are still growing. But NTES is not commercialize them. That represents approximately $16M of unrealized money. I wonder just how long is NTES going to wait before they start making money on these two games.

With the stock so high right now, I wonder what would happen if there is a revenue short fall in 1Q2009.

Medium to long term, I can’t express how strong NTES is going to be. There are significant activities in NTES’s 3G effort, portal effort, licensed games (WOW and Starcraft) and self developed games. I will have several write-up on each of these activities.

Two years from now, we might be looking at a company that is several times larger than the NTES of now.

NTES 4Q2008 Revenue Prediction Error

Based on my weekly tracking of NTES's game servers, I made a prediction of NTES's game revenue. You can find it here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/ntes-games-4q2008-prediction.html

Boy, was it a horrible prediction, talking about completely off the mark. I tabulate my prediction, compared with the actual result, in the following table:

	4Q08
Predicted Game Revenue (M)	110
Actual Game Revenue (M)	98.6
% Error	11.6%
Predicted XY2 ACU (k)	211
Actual XY2 ACU (k)	197
% Error	7.1%
Predicted XYQ ACU (k)	713
Actual XYQ ACU (k)	625
% Error	14%

Note: 62 of 65 days of summer holiday occured in 3Q08.

The above table shows that my prediction was absolutely horrible.

Every week, for 2 to 3 days a week, I got up early in the morning and late at night to track NTES's game server and obtain its statistics. It is an extremely time consuming work. It is very depressing that my final result is so far from the real data.

Going back, I think I understand the reason. The way I calculate the 4Q ACU (Average Concurrent User) is by comparing my measured statistics with the actual ACU from the 3Q. But by doing that, I no longer am able to take account the seasonal factor.

I can make year over year comparison, but that would introduce other problems. NTES increase its game server capacity from time to time. Therefore, the server statistics means different from that of a year ago.

There is a seasonal factor I need to account for. I need to figure out a factor to account for the summer holiday.

But for the mean time, I will no longer provide revenue predictions.

Sohu’s TLBB 4Q2008 Prediction

My last article on TLBB can be found here:

http://chinese-net-gaming-stock.blogspot.com/2008/07/sohu-2q2008-prediction.html

In this article, I am going to calculate the percent of growth for the game XYQ from 3Q 2008 to 4Q2008. I will be doing it using the server count.

The acronyms first -

ACU = Average Concurrent Users

PCU = Peak Concurrent Users

APA = Active Paying Account

ARPU = Average (Quarterly) Revenue per (Active Paying) User

QoQ = Quarter over Quarter

YoY = Year over Year

Now, let's do the 4Q2008 estimate for XYQ.

The server counts for the XYQ can be found in the following table:

There are 3 months in a quarter. Each month has either 4 or 5 weeks. I only use 2 days in a week. I don't do server count every weekend. I only put in the number where I have a server count in both quarter. I also pick 3 days of holidays in this calculation.

I have some computer problem during the months of November and December. That causes me to miss the server count for most of these two months. Therefore, I am not able to get the amount of server counts as I would like.

From the above table, I have a 4Q2008 ACU increase of 5.37% quarter over quarter (QoQ).

To calculate the revenue estimate for the 4Q2008, I use the following table:

	1Q2008	2Q2008	3Q2008	4Q2008
ACU QoQ		4.47%	-0.77%	5.37%
Revenue	38.9	45.5	51	58.65
Revenue QoQ	77%	17%	12%	15%
APA	1.39	1.68	1.86	2.14
APA QoQ	26.4%	21%	10.7%	15%
ARPU	199	179	178	178

The second row is the ACU QoQ increase (or decrease). SOHU only provides PCU in its earning conference. The ACU are my estimates and therefore is not official. The third row is quarterly revenue in millions of dollars. The fourth row is the QoQ comparison for the quarterly revenue. The fifth row is the APA in millions of players. The sixth row is the QoQ comparison for the APA. The Final row is the ARPU in Chinese currency (RMB).

Note that every number in the last column (for the 4Q2008) are my estimates. From previous table, I derive an ACU QoQ estimate of 5.37%. I use that number to give me a rough estimate for the APA QoQ. I conservatively estimate the 4Q APA QoQ number to be 15%. For ARPU, I think it has reach a stable point. Therefore, I use the last quarter estimate for the current quarter, or, it stays as 178 RMB.

With APA grows by 15% while ARPU stays the same, the revenue will grow by 15%. Therefore, the 4Q2008 revenue is estimated to be 58.65 millions.

Now I need to give derive an estimate for SOHU's other game, Blade Online (BO). The following table is the revenue for BO:

	4Q2007	1Q2008	2Q2008	3Q2008	4Q2008
BO Revenue	2.0	2.1	2.4	3.6	4

If any thing can symbolize how can the number 1 gaming vertical, 17173.com (owned by Sohu) and number 2 internet portal can help drive traffic to a game, the recent rejuvenation of BO speaks volumes.

Even though TLBB is not a revolutionary game nor is it very pretty, it is a fun game to play. But a game like BO is a game that really has no redeeming features to it. On top of that, it is a licensed game.

But in the end, a game like BO will not be a major game. I think the recent surge will level off. I think the level off will occur this quarter. Therefore, I predict the revenue of 4 millions dollars.

Adds up the revenue estimates for the two games, we get a quarterly gaming revenue of 62.65 millions for this quarter. It is significantly more than the guidance of 57 to 59 millions SOHU given in the 3Q earning conference.

But I don't have much visibility on how well (or badly) SOHU's other branches (Advertising and wireless) will do this quarter.

NTES Games 4Q2008 Prediction

This is the final part of a long series of articles.

For the article on XYQ, it can be found here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/xyq-4q2008-prediction.html

For the article on XY2, it can be found here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/xy2-4q2008-prediction.html

For the article on XY3, it can be found here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/xy3-4q2008-prediction.html

For the article on XYW, it can be found here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/xyw-4q2008-prediction.html

For the article on TX2, it can be found here:

http://chinese-net-gaming-stock.blogspot.com/2009/01/tx2-4q2008-prediction.html

Now I have the revenue prediction for TX2 and XYW. I also have the ACU prediction for XY2, XY3, and XYQ. Later on in this article, I will convert the ACU into the revenue estimates.

But first, I need to talk about DT, FF2, and DT2.

DT is a failed game. Its revenue per quarter had been fluctuate around 0.6 million to 0.8 million. I will use an estimate of 0.7 millions of revenue for DT this quarter.

DT2 is still in closed beta testing as of now. There will be no revenue from this game in this quarter.

FF2 is in advanced stages of closed beta testing. It will most likely be released in the 1Q2009. But as of now, I don't think NTES is getting any revenue from this game yet.

XYQ is doing very well. It is still growing at a very healthy clip.

XY2 is a big surprise. An ACU QoQ increase of 24% is simply amazing. XY2 is just a breath away from being the second million players mega-hit game for NTES.

XY3 is disappointing. Its traffic actually decreased in this quarter.

But XYW is the new star for this quarter. It has mega-hit written all over it. Even though its revenue contribution is very small during this quarter. But that is to be expected for a free-to-play game. Unlike paying games (where ACU and revenue have a linear relationships), a free-to-play game tends to explode exponentially during the first few quarters (from very low level of revenue in the first quarter to very high level in the next few quarters). XYW has all the marking to be the next mega-game after XYQ and XY2 for NTES. Most of all, I don't think NTES expected that, certainly not me.

XY2 has a very long lineage. It is one of the very rare mega game in China. Then NTES modified it and make a more cartoon-like game out of it in the form of XYQ. XYQ eventually becomes the number 1 game in China (some say in the world).

As XY2 getting long in the tooth. NTES builds its sequel, XY3, from the ground up. But somehow, XY3 is not as successful as NTES expected.

NTES made the decision to operate both games, XY2 and XY3, at the same time. But surprisingly, it is the older game (XY2) that starts to grow in an exponential way while the newer game languish.

NTES spent a lot of resource to develop the underlining game engine for XY3. NTES wants to get more out of its investment than just XY3. So they build the XYW. To everyone's surprise, XYW took off like a rocket.

Currently, NTES has four lines of games in its arsenals. The XY2 line of games will probably have 3 mega-hit games out of it. All four games are 2D turned based Cartoon style MMORPG game. XYQ is more cartoon-like while XY2 has more realistic cartoon drawing. XY3 is the same as XY2 except it is build on a brand new game engine. XYW is basically the same as XY2 and XY3 except it is the only free-to-play game out of the four.

The FF1 game is a failure so far. Hopefully FF2 will do better. FF1 and FF2 are cartoon like MMORPG games. They are free-to-play games. They are the only western themed games for NTES. They are also the only modern theme games for NTES. I expect FF2 will be released in 1Q (2Q at the latest) this year.

The DT1 game is also a failure. Again, hopefully DT2 will do better. The DTs are the so called 2.5D MMORPG games. They are highly realistic and very beautiful to behold. The underlining game engine was written in-house. DT1 is a paying game while DT2 will most likely be a free-to-play game. I think DT2 will be ready to be released by the 2Q2009. But there is a possibility that NTES will delay that for marketing reasons.

The fourth line of game is TX2. So far it is on its way to be a major game. Whether it be another mega game (half of million or more of players) is hard to tell. TX2 is NTES's first venture into the 3D MMORPG games. It is a free-to-play game. The game engine is licensed from an Australian company. Part of the reason of why this game might be hard to become a mega-hit game is because it requires users to have powerful computers to play this game. But I do think the revenue per users will be very high.

Now, let's find an estimate for the revenue for XY2, XY3, and XYQ. I will use the following table to calculate the estimate:

	1Q2008	2Q2008	3Q2008	4Q2008
XY2 ACU	152	149	170	211
XY3 ACU	46	51	48	48
XYQ ACU	583	627	675	713
Total ACU	781	827	893	972
Game Revenue	79.3	86.8	99.4
Revenue - DT	78.6	86.1	98.7	104b
(Rev.-DT)/ACU	0.1	0.104	0.11	0.107a

The 2^nd to 4^th row are the ACU number for the three games. When you adds up all three rows, you get the 5^th row, total ACU. The 6^th row is the game revenue. Assuming DT has revenue of 0.7 million per quarter, the 7^th row is the total game revenue – DT revenue. By dividing that number by the total ACU, it will give us the 8^th row.

From 2Q2008 to 3Q2008, (Rev –DT)/ACU ranges from 0.104 to 0.11. For the 4Q2008, I pick the middle number (0.107). By multiplying 0.107 with the total ACU of 972k, I got a Revenue-DT = 104 millions.

Now, we can find the total gaming revenue for NTES. It is calculated in the following table:

	Revenue
XY2 + XY3 + XYQ	104
DT	0.7
XYW	0.7
TX2	4.6
Total	110

Finally, we got the total gaming revenue of 110 million for the 4Q2008. Now, to get the total quarterly revenue estimate for the whole company, all one has to do is to add the estimated revenues for advertising and wireless services.

TX2 4Q2008 Prediction

My previous article on TX2 can be found here:

http://chinese-net-gaming-stock.blogspot.com/2008/09/ntes-tx2-status-part-2-official-closed.html

Tx2 is a new game. It starts commercial operation on 6/6/2008. It also had a new expansion pact on 10/29/2008. Therefore, one shall expect strong growth in the next few quarters. If the game is a mega hit, one shall expect strong growth for the next few years.

The acronyms first -

ACU = Average Concurrent Users

PCU = Peak Concurrent Users

QoQ = Quarter over Quarter

YoY = Year over Year

Now let us see how TX2 does in 4Q2008.

The server counts for the XY2 can be found in the following table:

There are 3 months in a quarter. Each month has either 4 or 5 weeks. I only use 2 days in a week. I don't do server count every weekend. I only put in the number where I have a server count in both quarter. I also pick 3 days of holidays in this calculation.

The 31.2% QoQ increase in ACU is very nice. NTES doesn't give out statistics for this game. Therefore, I don't know the PCU record for this game. Note that NTES doesn't have a company guideline on which information to publish, it depends on the each game's chief designer.

But I best guess is TX2 already has PCU of a minimal of 100k. Considering this game is only half years old, I full expect this game will grow to a major game of a quarter of million players (at the minimal). In the best case, it is not out of question that it could eventually achieve the popularity of World of Warcraft.

TX2』s chief designer doesn't give out ACU estimates also. My estimate is an ACU = 40k for 4Q2008.

TX2 is a free-to-play game. Therefore, I am going to calculate its revenue for this quarter. Again, NTES doesn't provide any historical revenue data for TX2. I will have to use other free-to-play game as guideline.

I will use Sohu's TLBB as a guide. During the 3 quarter in operation, TLBB had achieved a PCU of 500k while generated 23 million in revenue. This quarter (4Q2008) is at the similar stage in operation, the third quarter in operation. TX2 is a 100k game in this quarter. Therefore, my estimate for revenue is 23 * (100k/500k) = 4.6 millions.

By the way, if case anybody is wondering why is Sohu's TLBB can achieve a PCU of 500k while TX2 can only achieve a PCU of 100k at roughly the same period of their respective game life, the discrepancy is largely because Sohu really marketing its game while NTES tends to pinch pennies and tries to save the marketing cost. As a result, TLBB PCU reached its plateau right away while TX2 keeps on explosive growing.