Evaluation of citrus market price risk based on kernel density estimation

Taking the monthly wholesale prices of 4 kinds of citrus（Sweet orange ，Ponkan，Mandarin， Naval orange） from 2014 to 2016 as the object of study，the non-parametric kernel density estimation method was used to establish a model to evaluate price risk. The results showed that，different kinds of citrus had different seasonal characters in price. The price of Mandarin arrived at peak during August to September，the wave pattern was inverted“ V” type，which was low at both sides and high in the middle. And the price of Naval orange had two peaks during July to August and the wave pattern was“ M” type. The descending order of fluctuating intensity of citrus was Mandarin，Ponkan，Sweet orange，Naval orange，but there was a little difference among four kinds of citrus. Above all，the risk of citrus market was high，and the probability of declining was higher than rising in price. Among the four kinds of citrus，the highest risk appeared in Mandarin，followed by Ponkan，and the probability of declining in Sweet orange price was similar to Naval orange’s. Overall，the price risk of Naval orange was the lowest among the four citrus.