**Summary:** In this article, I forecast China's Q3 GDP growth. I define possible scenarios for Q3 growth based upon recent news, map these scenarios to past growth, construct a predictive model, and forecast Q3 GDP growth based upon history, the model, and my views.

**Define possible scenarios:** In this section, I define Best-Case and Worst-Case scenarios for GDP growth, suggest an interval for Q3 Consensus estimates, and interpolate Above-Consensus and Below-Consensus scenarios.

On July 23, 2013, Premier Li Keqiang set a floor of 7% for annual GDP growth. GDP growth below this figure would undermine the credibility of the government. Therefore, I define this to be the Worst-Case scenario. In addition, China's GDP has increased at less than an 8% annual growth rate over the past two years. Consequently, I set GDP growth of 8% or above as the Best-Case scenario.

*Figure 1 shows the annual growth rate of China's real GDP ((adjusted for inflation)) at a quarterly frequency from Q1 2002 to Q1 2014.*

*Source: National Bureau of Statistics of China and author's calculations*

Next, I suggest the Consensus interval for Q3 GDP growth will be between 7.4% and 7.6%, based on the following considerations. China's first-quarter GDP increased 7.4% over last year, and both Nomura and Barclays expect 7.4% in the second quarter. This is slightly higher than the 7.3% consensus estimate from Reuters in April. More recently, Li Keqiang stated he is willing to adjust policy to ensure GDP growth of 7.5% in 2014. Given Q1 and consensus Q2 growth, China will need to grow faster than 7.5% in the second half of the year to match this target. Therefore, I anticipate consensus Q3 growth of between 7.4% and 7.6%.

Lastly, I define Below-Consensus to be the interval between the Worst-Case scenario and Consensus: 7% to 7.4%. Similarly, Above-Consensus is the interval between the Best-Case scenario and Consensus: 7.6% to 8%.

Figure 2 summarizes these scenarios and the corresponding intervals. I will forecast GDP in terms of probabilities and expectations based upon these intervals.

**Figure 2 maps each scenario to an interval of China's GDP growth. For example, the Worst-Case scenario corresponds to GDP growth less than 7%.**

*Source: Author*

**Map Scenarios to History:** Next, I map the above scenarios to historical GDP growth and show they align with conservative assumptions from a historical perspective. I sourced data from the National Bureau of Statistics of China (NBSC), verified my calculations against World Bank figures, deflated GDP to year-2000 Yuan, and calculated annual GDP at a quarterly frequency. First, I introduce standard components of GDP and decompose these into Fast Growth and Slow Growth categories. Then, I suggest low, medium, and high estimates for both Fast Growth and Slow Growth, and connect these with the aforementioned intervals.

NBSC reports the main components of GDP as Primary, Secondary, and Tertiary sectors. These roughly correspond to the Agricultural, Industrial, and Service sectors. Figure 3 shows the Service sector is increasing at a faster rate than either Agriculture or Industry, and it surpassed the Industrial sector in the first quarter of 2013.

**Figure 3 illustrates the value of the three components of China's GDP from Q1 2002 to Q1 2014. All values are in year-2000 Yuan.**

*Source: National Bureau of Statistics of China and author's calculations*

Next, I split these GDP categories into Fast Growth and Slow Growth. Fast Growth consists of the Service industries that consistently grew faster than 10%, including the Financial, Wholesale and Retail, and Other Lines ((a miscellaneous category)) industries; whereas, Slow Growth contains the Real Estate, Transportation, and Accommodation service industries. The Agricultural and Industrial sectors account for the remainder of Slow Growth.

Figures 4a and 4b show Slow Growth is twice the weight of Fast Growth as a percentage of GDP, and Fast Growth increased at double the rate of Slow Growth. Assuming these growth rates stay constant, Slow Growth and Fast Growth should contribute approximately equally to aggregate GDP growth.

**Figure 4a shows that Fast Growth GDP increased from slightly less than half of Slow Growth GDP by weight to slightly more between 2012 and 2014. Figure 4b illustrates that the growth rates of Fast Growth and Slow Growth were stable over the same period, and Fast Growth increased at twice the rate of Slow Growth.**

*Source: National Bureau of Statistics of China and author's calculations*

Figure 4b and Figure 5 outline potential scenarios for aggregate Q3 GDP growth, based on low, medium, and high estimates of Fast Growth and Slow Growth. Each of these aggregate scenarios corresponds to an interval from section 1. Although the low, medium, and high estimates for Fast Growth align with history, those for Slow Growth are below historical values. In effect, I assume that 2014 Q1 for Slow Growth is indicative of the future. Therefore, the intervals enumerated in section 1 correspond with a relatively conservative view of future GDP growth.

*Figure 5 displays low, medium, and high estimates for Slow Growth and Fast Growth Q3 GDP growth. As shown in Figure 4b, the low, medium, and high estimates for Fast Growth roughly correspond to the historical minimum, average, and maximum growth rates. The distance between the medium and other two estimates is .8%. In contrast, the Slow Growth medium value approximately matches the 2014 Q1 value, and the low and high values are .4% from medium, or half the distance of the Fast Growth cases. Each estimate for GDP growth maps to the aforementioned intervals in section 1, as follows: red = Worst-Case, orange = Below-Consensus; yellow = Consensus; light green = Above-Consensus; and green = Best-Case.*

*Source: Author*

**Constructing a predictive model:** In this section, I build a predictive model of GDP growth. The model indicates Q3 growth is likely to match or exceed Consensus.

I constructed a few models based on annual GDP growth data ((y)) from the NBSC and loan growth data ((x)) from China Construction Bank (OTCPK:CICHF) Interim and Annual Reports from 2005 to 2013. In effect, I use the China Construction Bank data as a proxy for total loan growth in China's banking sector. To estimate the models, I used semiannual data to estimate the coefficients in the following equation:

In words, GDP growth this year is a linear combination of GDP growth last year and loan growth last year. Next, I estimated the above model using Total, Corporate, Personal, and Real Estate loan growth. All of these variables are statistically significant predictors of GDP growth. The Real Estate loan model explained the most variance ((77% of future GDP variance)). Figure 6 shows the results of this regression.

**Figure 6 illustrates the results of a multiple regression with GDP this year [GDP((t))] as the dependent variable and GDP last year [GDP((t-2))] and % change in Real Estate loans last year [RE((t-2))] as the independent variables. Both variables are significant, at a 95% confidence level ((TSTAT > 2.2)).**

*Source: National Bureau of Statistics of China, 2005 through 2013, Interim and Annual Reports for China Construction Bank, and author's calculations*

Figure 7 compares predicted GDP growth from the Real Estate model to actual GDP growth. The forecast matches the trend of actual GDP fairly well; although, predicted GDP growth has exceeded actual GDP growth for the last two and a half years. Therefore, inferences should be drawn with caution. Nevertheless, the model predicts GDP growth will trend upwards over the second half of the year. This suggests Q3 GDP growth will equal or exceed Consensus.

*Figure 7 compares forecast to actual real GDP growth.*

*Source: National Bureau of Statistics of China, Interim and Annual Reports for China Construction Bank from 2005 to 2013, and author's calculations*

**Estimate probabilities associated with each scenario:** In this section, I base forecasts for the probabilities and expected values of each interval on the historical analysis, predictive model, and qualitative considerations outlined in the previous three sections. I conclude Q3 GDP should match or exceed Consensus.

The historical analysis, predictive model, and comments of China's Premier all indicate Q3 GDP growth will exceed 7.4%. However, I expect the increase over 2014 Q1 will be slight. Therefore, I conclude the Consensus scenario is the most likely, and assign it a probability of 45%. Using similar logic, I anticipate Above-Consensus is more likely than Below-Consensus. Accordingly, I assign probabilities of 25% and 20% to Above-Consensus and Below-Consensus, respectively.

Next, I focus on the tails. Growth of less than 7% would suggest China's government has both lost control over the economy and reports GDP accurately. I believe the joint probability of these occurring is low. To allow for a factor of safety in the estimates, I assign Worst-Case a probability of 2%, or 1 in 50. Although reported GDP growth above 8% is also unlikely, it seems more likely than the Worst-Case scenario. Therefore, I associate a 3% probability with the Best-Case scenario.

Figure 8 summarizes my forecast for China's GDP.

*Figure 8 shows the forecast probabilities associated with five possible intervals for China's GDP growth. For example, I forecast a 3% chance of GDP exceeding 8%.*

*Source: Author*

Figure 9 represents my expectations for the outcome within each interval. Within each interval, I expect the outcome to be biased towards Consensus, on average. In the Worst-Case and Best-Case scenarios, this translates to expectations .1% beneath and above the cut-off, respectively. For Below-Consensus and Above-Consensus, I set the expected GDP to be .1% removed from Consensus. Lastly, I expect GDP for the Consensus will equal the Premier's targeted growth of 7.5%.

**Figure 9 illustrates the expected GDP within each interval. For example, if the GDP in Q3 is less than 7%, then I expect it to be approximately 6.9%.**

*Source: Author*

**Disclosure: **The author has no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.