Eden Rabbie, September 2020

30-second summary

- UAE startups with access to capital at series A and above create more jobs than unfunded startups across the board
- UAE startups who lack of access to capital are 1.5 times more likely to go out of business
- The odds of a UAE startup of raising a startup round, compared to startups with prior access to capital are 1:3
- Funded UAE startups end up producing an economic value similar to traditional SMEs, and they both produce exits of similar value

**Intrigued, but busy?**

Skip to the Final Remarks

Despite reports in the media showing growth in deployed venture capital and loan options for UAE startups, access to capital continues to be repeatedly reported as a key challenge for UAE startups, highlighted even by the former director of IMF. And as Covid-19 hit in 2020, access to capital is reported to be even more challenging.

**So what?**

Is there a basis more concrete than common sense and self-reporting surveys that would compel us to care about improving the slowing situation of UAE startup access to capital?

This report answers this question and outlines the opportunity cost. We analyze 1,563 startups from UAE and MENA, as well as 1,527 SMEs from UAE, to quantify the economic impact which access to capital may have on UAE startups and their contribution to the economy with scientific reliability.

- Does access to capital affect the number of jobs created by UAE startups
- Does lack of access to capital affect startup’s chances of going out of business in UAE
- Does prior access to capital affect startup’s chances of raising new funding in UAE
- Does access to capital affect startup’s chances of exiting in UAE
- Final remarks

- Does access to capital affect the number of jobs created by UAE startups
- Does lack of access to capital affect startup’s chances of going out of business in UAE
- Does prior access to capital affect startup’s chances of raising new funding in UAE
- Does access to capital affect startup’s chances of exiting in UAE
- Final remarks

2. A – Jobs

Comparison: Funded UAE startups, Unfunded UAE startups, Non-growth-type UAE SMEs who exited

2. A. 1 – Seed has no significant impact on jobs

- There is no significant difference in the number of jobs created by funded UAE startups at seed stage and unfunded ones.<sup>20</sup>
- There is also no significant difference in the number of jobs created by funded UAE startups at seed stage and non-growth-type SMEs.</sup>21,†</sup>

2. A. 2 – Positive impact of series A and above on jobs

**Funded UAE startups at series A are 1.8 times more likely to create more than 10 jobs compared to unfunded UAE startups.**<sup>22</sup> At least 75% of UAE startups at series A have more than 10 employees.<sup>23</sup>**Funded startups at series B and above are 2.2 times more likely to create more than 50 jobs compared to startups who exit without prior funding**.<sup>24</sup>

2. A. 3 – More funding, more jobs

**Funded UAE startups at series A stage are 2.1 times more likely to create more than 10 jobs compared to seed stage.**<sup>25</sup>**Funded UAE startups at series B stage and above are 4.7 times more likely to create more than 50 jobs compared to series A stage.**<sup>26</sup>

2. A. 4

**Funded UAE startups at series B stage and funded exits don’t create more jobs when compared to non-growth-type UAE SMEs who exit.**<sup>27</sup>- This is a curious finding which we will address later in this report series.

Technical notes (8 notes, 20-27)

<sup>Note 20:</sup> Difference of count of funded UAE startups at seed stage vs unfunded UAE startups, both reporting the event of having over 10 employees, is not statistically significant. X<sup>2</sup>(1, N=432)=1.48, p=.22.

<sup>Note 21:</sup> Difference of count of funded UAE startups at seed stage vs non-growth-type UAE SMEs, both reporting the event of having over 10 employees, is not statistically significant. X<sup>2</sup>(1, N=2505)=1.54, p=.22.

<sup>Note 22:</sup> The effect of series A funding on creating over 10 jobs reported as relative risk of number of funded UAE startups at series A stage vs unfunded UAE startups, both reporting the event of having >10 employees, is statistically significant s.t. RR lower bound > 1. RR=1.8, 95% CI [1.52, 2.13] (statistically significant); N=371, p<sub>2T</sub>=.001 Fisher Exact Test (correlation is statistically significant).

<sup>Note 23:</sup> Population proportion of startups who reported having over 10 employees out of total funded UAE startups at series A stage, p=.93. N=17, 95% CI [.75, 1].

<sup>Note 24:</sup> The effect of series B funding on creating over 50 jobs reported as relative risk of number of funded UAE startups at series B stage vs at unfunded UAE startups who had exited, both reporting the event of having over 50 employees, is statistically significant s.t. RR lower bound> 1. RR=2.21, 90% CI [1.04, 4.69]; N=33, p<sub>2T</sub>=.0799 Fisher Exact Test (correlation is statistically significant with alpha at .10).

<sup>Note 25:</sup> The effect of series A funding on creating over 10 jobs reported as relative risk of number of funded UAE startups at series A stage vs at seed stage, both reporting the event of having over 10 employees, is statistically significant s.t. RR lower bound > 1. RR=2.14, 95% CI [1.57, 2.93] (statistically significant); N=77, p<sub>2T</sub>=<.001 Fisher Exact Test (correlation is statistically significant).

<sup>Note 26:</sup> The effect of series B funding on creating over 50 jobs reported as relative risk of number of funded UAE startups at series B stage vs at series A stage, both reporting the event of having over 50 employees, is statistically significant s.t. RR lower bound > 1. RR=4.74, 95% CI [1.54, 14.52] (statistically significant); N=34, p<sub>2T</sub>=.0051 Fisher Exact Test (correlation is statistically significant).

<sup>Note 27:</sup> The effect of series B funding on creating more jobs reported as relative risk of number of funded UAE startups at series B stage vs non-growth-type UAE SMEs who had exited, both reporting the event of having over 50, over 100, and over 250 employees, is not statistically significant in each case s.t. RR lower bound ≤ 1. For the 3 events: over 50 employees, over 100, and over 250, respectively, RR 95% CI ([0.66, 1.52], [0.38, 1.44], [0.34, 2.02]); p<sub>2T</sub>=(1.00, .35, .49).

Number of employees is the most recent number self-reported by the startup, processed as nominal groups: 1-10, 11-50, 51-100, 101-250, >250 .

Startup stage is attributed based on its most recent closed round.

Correlation is studied through Pearson’s chi-squared test iff. each observation o ≥ 5, and Fisher Exact Test otherwise.

Statistical significance is studied through analysis of p-value and relative risk confidence interval.

Effect size is studied through analysis of relative risk confidence interval.

Unless explicitly mentioned, alpha = .05.

Null value for relative risk is 1.

N is startups aged 2+ years, measured as founded on or before 31 Dec 2017 to allow for median age to raise first startup round to take effect.

For Note 23: Statistical significance and effect size are studied via analysis of confidence interval. CI is Wilson score interval.

† Non-growth-type SME is a business that is not product-based, not technology R&D based, not aiming at disrupting a technology support net, and not growth stock. This group includes businesses such as general trading companies, restaurants, consulting agencies, software development boutiques and marketing agencies.

Source

Rabbie, Eden. (2020). UAE Startup Access To Capital 2010-2020. *Startup Report MENA* [Online]. https://startupreport.me/uae-startup-access-to-capital-2020 (Based on analysis of 1011 equity-only seed, series A, bridge and series B startup rounds, 443 investors, 4730 ventures in MENA 2008-2019; enriched database: Crunchbase, PitchBook and fieldwork. Data last accessed 30 April 2020).

Armitage, P., Berry, G. and Matthews, J. N. S. (2001). *Statistical Methods in Medical Research*. 4th ed. Hoboken, NJ: Wiley-Blackwell. ISBN 978-0-632-05257-8.

Newcombe, R. G. (1998). Interval estimation for the difference between independent proportions: comparison of eleven methods. *Statistics in medicine*, 17: 873-890. doi:10.1002/(SICI)1097-0258(19980430)17:8<873::AID-SIM779>3.0.CO;2-I

Snedecor, G. W. and Cochran, W. G. (1989). S*tatistical Methods*. 8th ed. Ames, IA: Iowa State University Press. ISBN 978-0813815619.

Wallis, S. (2013). Binomial Confidence Intervals and Contingency Tests: Mathematical Fundamentals and the Evaluation of Alternative Methods. *Journal of Quantitative Linguistics*, 20:3, 178-208. doi:10.1080/09296174.2013.799918

2. B – Survival

UAE startups who lack access to capital are 1.5 times more likely to go inactive<sup>28</sup>

2. B – Survival

- This is similar to the rest of MENA region.<sup>29</sup>

Technical notes (2 notes, 28-29)

<sup>Note 28:</sup> The effect of lack of access to capital on going inactive reported as relative risk of number of unfunded vs funded UAE startups, both reporting the event of being inactive, is statistically significant s.t. RR lower bound > 1. RR=1.45, 95% CI [1.03, 2.05] (statistically significant); X<sup>2</sup>(1, N=701)=4.73, p=.029 (correlation is statistically significant).

<sup>Note 29:</sup> Difference in the effect of lack of access to capital on going inactive reported as the difference of relative risk of number of unfunded vs funded startups, both reporting the event of being inactive, in UAE vs non-UAE MENA, is not statistically significant. N<sub>pooled</sub>=1563, 95% CI [-.63, .67]. For UAE relative risk, see Note 28. Non-UAE MENA relative risk RR=1.47, 95% CI [1.13, 1.91] (statistically significant); X<sup>2</sup>(1, N=862)=8.49, p=.036 (correlation is statistically significant).

Inactivity is observed through our model which uses for input self-reported closures, website HTTP status code and job data on the startup’s co-founders LinkedIn profiles to classify inactivity status. Model accuracy=.80, precision=.73, sensitivity=.88.

Correlation is studied through Pearson’s chi-squared test iff. each observation o ≥ 5, and Fisher Exact Test otherwise.

Statistical significance is studied through analysis of p-value and relative risk confidence interval.

Effect size is studied through analysis of relative risk confidence interval.

Unless explicitly mentioned, alpha = .05.

Null value for relative risk is 1.

N is startups aged 2+ years, measured as founded on or before 31 Dec 2017 to allow for median age to raise first startup round to take effect.

For Note 29: Statistical significance and effect size are studied via analysis of relative risk confidence interval. CI is Wilson score interval.

Source

Rabbie, Eden. (2020). UAE Startup Access To Capital 2010-2020. *Startup Report MENA* [Online]. https://startupreport.me/uae-startup-access-to-capital-2020 (Based on analysis of 1011 equity-only seed, series A, bridge and series B startup rounds, 443 investors, 4730 ventures in MENA 2008-2019; enriched database: Crunchbase, PitchBook and fieldwork. Data last accessed 30 April 2020).

Armitage, P., Berry, G. and Matthews, J. N. S. (2001). *Statistical Methods in Medical Research*. 4th ed. Hoboken, NJ: Wiley-Blackwell. ISBN 978-0-632-05257-8.

Newcombe, R. G. (1998). Interval estimation for the difference between independent proportions: comparison of eleven methods. *Statistics in medicine*, 17: 873-890. doi:10.1002/(SICI)1097-0258(19980430)17:8<873::AID-SIM779>3.0.CO;2-I

Snedecor, G. W. and Cochran, W. G. (1989). S*tatistical Methods*. 8th ed. Ames, IA: Iowa State University Press. ISBN 978-0813815619.

Wallis, S. (2013). Binomial Confidence Intervals and Contingency Tests: Mathematical Fundamentals and the Evaluation of Alternative Methods. *Journal of Quantitative Linguistics*, 20:3, 178-208. doi:10.1080/09296174.2013.799918

2. C – Access to more capital

**1:3 odds**: Compared to a UAE startup who had raised a startup round before, the likelihood of a UAE startup raising their first startup round is 61% lower<sup>30</sup> (1:3 odds).<sup>31</sup>**These odds remain at the same level since 2017.<sup>32</sup>**The odds of a UAE startup raising their first startup round had improved significantly 3 years ago, from 1:8 in 2014 to 1:3 in 2017, compared to a UAE startup who had raised a startup round before.<sup>33</sup> It remains the same ever since.<sup>32</sup>**This is different from the rest of MENA region.**Since 2016 both new startups and startups with prior funding have had similar odds of raising a startup round.<sup>34</sup>

Technical notes (5 notes, 30-34)

<sup>Note 30:</sup> The effect of lack of access to capital on successfully raising a startup round reported as relative risk of number of unfunded vs funded UAE startups, both reporting the event of raising one more startup round, is statistically significant s.t. RR upper bound < 1. RR=0.61, 95% CI [0.46, 0.81] (statistically significant); X<sup>2</sup>(1, N=811)=37.57, p<.001 (correlation is statistically significant).

<sup>Note 31:</sup> The effect of prior access to capital on successfully raising a startup round reported as odds ratio of number of funded vs unfunded UAE startups, both reporting the event of raising one more startup round is > 1:1 s.t. OR lower bound > 1. OR=3.6, 95% CI [2.4, 5.6] ; X<sup>2</sup>(1, N=811)=37.57, p<.001 (correlation is statistically significant).

<sup>Note 32:</sup> Change in the effect of prior access to capital on successfully raising a startup round reported as difference in odds ratio of number of funded vs unfunded UAE startups, both reporting the event of raising one more startup round, with 5-year window, measured on 2019 vs 2017 and on 2019 vs 2018, is not statistically significant in either pairs. 2019 vs 2017 and 2019 vs 2018, respectively: N<sub>pooled</sub>=(1144, 952), 95% CI ([-1.73, 2.38], [-2.09, 2.35]). Odds ratio on 2017, 2018 and 2019, respectively, OR=(2.9, 2.6, 2.7), 95% CI ([1.7, 4.7], [1.6, 4.2], [1.5, 4.7]); X<sup>2</sup>((1, N=607)=18.11, (1, N=537)=14.19, (1, N=415)=12.35), p<.001 in each case (correlation is statistically significant).

<sup>Note 33:</sup> Change in the effect of prior access to capital on successfully raising a startup round reported as difference in odds ratio of number of funded vs unfunded UAE startups, both reporting the event of raising one more startup round, with 5-year window, measured on 2017 vs 2014, is negative. δ=-5.18, N<sub>pooled</sub>=619, 95% CI [-16.40, -0.00005] (statistically significant). Odds ratio on 2014, OR=7.8, 95% CI [3.2, 19.2]; X<sup>2</sup>((1, N=204)=25.06, p<.001 (correlation is statistically significant). For 2017, see Note 31.

<sup>Note 34:</sup> The effect of prior access to capital on successfully raising a startup round reported as relative risk of number of funded vs unfunded non-UAE MENA startups, both reporting the event of raising one more startup round, with 5-year window, measured on 2016,.., 2019, is not statistically significant in either year. RR 95% CI ([0.93, 1.91], [0.78, 1.51], [0.72, 1.44], [0.63, 1.35]); N=(640, 752, 825, 762), p=(.129, .646, .928, .680) (correlation is not statistically significant).

Correlation is studied through Pearson’s chi-squared test.

Statistical significance is studied through analysis of p-value and relative risk confidence interval.

Effect size is studied through analysis of relative risk confidence interval and odds ratio confidence interval.

Unless explicitly mentioned, alpha = .05.

Null value for relative risk is 1.

N is startups aged 2+ years, measured as founded on or before 31 Dec 2017 to allow for median age to raise first startup round to take effect.

For Notes 32 and 33: Statistical significance and effect size are studied via analysis of difference in odds ratio confidence intervals. CI is Wilson score interval.

Source

Rabbie, Eden. (2020). UAE Startup Access To Capital 2010-2020. *Startup Report MENA* [Online]. https://startupreport.me/uae-startup-access-to-capital-2020 (Based on analysis of 1011 equity-only seed, series A, bridge and series B startup rounds, 443 investors, 4730 ventures in MENA 2008-2019; enriched database: Crunchbase, PitchBook and fieldwork. Data last accessed 30 April 2020).

Armitage, P., Berry, G. and Matthews, J. N. S. (2001). *Statistical Methods in Medical Research*. 4th ed. Hoboken, NJ: Wiley-Blackwell. ISBN 978-0-632-05257-8.

Newcombe, R. G. (1998). Interval estimation for the difference between independent proportions: comparison of eleven methods. *Statistics in medicine*, 17: 873-890. doi:10.1002/(SICI)1097-0258(19980430)17:8<873::AID-SIM779>3.0.CO;2-I

Snedecor, G. W. and Cochran, W. G. (1989). S*tatistical Methods*. 8th ed. Ames, IA: Iowa State University Press. ISBN 978-0813815619.

Wallis, S. (2013). Binomial Confidence Intervals and Contingency Tests: Mathematical Fundamentals and the Evaluation of Alternative Methods. *Journal of Quantitative Linguistics*, 20:3, 178-208. doi:10.1080/09296174.2013.799918

2. D – Exiting

**For UAE startups, there is no significant association between startup’s fundraising activity and exiting.**<sup>35</sup>**This is similar to the rest of MENA region**.<sup>36</sup>

Technical notes (2 notes, 35-36)

<sup>Note 35:</sup> The effect of access to capital on exiting reported as relative risk of number of unfunded vs funded UAE startups, both reporting an exit event, is not statistically significant. RR 95% CI [0.40, 2.16], N=340, p=.872 (correlation is not statistically significant).

<sup>Note 36:</sup> The effect of access to capital on exiting reported as relative risk of number of unfunded vs funded non-UAE MENA startups, both reporting an exit event, is not statistically significant. RR 95% CI [0.13, 1.09], N=441, p<sub>2T</sub>=.067 Fisher Exact Test (correlation is not statistically significant).

Correlation is studied through Pearson’s chi-squared test iff. each observation o ≥ 5, and Fisher Exact Test otherwise.

Effect size is studied through analysis of relative risk confidence interval.

Unless explicitly mentioned, alpha = .05.

Null value for relative risk is 1.

N is startups aged 5+ years, measured as founded on or before 31 Dec 2014 to allow for median age to exit to take effect.

Source

*Startup Report MENA* [Online]. https://startupreport.me/uae-startup-access-to-capital-2020 (Based on analysis of 1011 equity-only seed, series A, bridge and series B startup rounds, 443 investors, 4730 ventures in MENA 2008-2019; enriched database: Crunchbase, PitchBook and fieldwork. Data last accessed 30 April 2020).

*Statistical Methods in Medical Research*. 4th ed. Hoboken, NJ: Wiley-Blackwell. ISBN 978-0-632-05257-8.

Snedecor, G. W. and Cochran, W. G. (1989). *Statistical Methods*. 8th ed. Ames, IA: Iowa State University Press. ISBN 978-0813815619.

Impact of Access To Capital On UAE Startups – Final Remarks

**Jobs:** In terms of direct contribution to UAE economy, seed capital does not have significant impact on job creation. However, funded startups at series A and above create more jobs than unfunded startups across the board.

** **

**Survival:** Beyond gut feeling and self-reported surveys, UAE startups who lack of access to capital are 1.5 times more likely to go out of business. So are startups from the rest of MENA.

**Underwhelming exits: **It is noteworthy that UAE startups who had exited on M&A deals without raising any startup rounds constitute the majority of exits in UAE. This unfunded status may explain their underwhelming contribution to job creation.

**Startups morphing into SMEs:** When a non-growth-type traditional SME exits in UAE, it create as many jobs as startups at series B and above. This could mean that the as-of-yet nascent startup space in UAE has matured enough to reach parity with the tried-and-proven traditional SME model. It could also mean that both enterprise types follow similar practices in terms of management and scaling. This case of “startups morphing into SMEs” adds to the need for an intervention to establish a separate track of support to enable startups of qualitatively-different management that can result in more effective scaling.

**Unnecessary risk:** It can be argued that if well-funded UAE startups end up producing an economic value similar to traditional SMEs, then investors are justified to move away from the higher risk UAE startup model. Indeed, investors seem to prefer avoiding investing in new startups in UAE. The odds of a new startup raising a startup round compared to startups with prior access to capital are 1:3. This improved significantly in 2017, showing a change in investors behavior at the time. However, it has not changed since 2017, signalling a return to the risk-averse behavior towards new startups when it comes to funding.

All this aligns with the slowdown of startups access to capital, which data shows in Insight 1c. (See Report 1: Situation By 2020)

In the next report in this series, we explore startup investor behavior in UAE to confirm if investors have been avoiding UAE startups, and if so, why.

Data Analytics and Commercialization in MENA

– 12 years of experience working with governments developing their SME sectors.

– Data science background, CIM, Six Sigma. Japanese-trained.

– Featured author on Medium.

Worked 8 years with the Japanese government to help MENA governments develop their SME sectors, followed by 3 years with UAE government to grow its venture space.

Copyright © startupreport.me and Eden Rabbie 2020-2023. Startup Report MENA is operated by __Clearworld__, PO Box 415816 Dubai, UAE.

You may use any part of this report for any purpose on condition that you link to it and cite it properly.

Rabbie, Eden. (2020). UAE Startup Access To Capital 2010-2020. Startup Report MENA [Online]. https://startupreport.me/uae-startup-access-to-capital-2020

Do you have information that can make this report more accurate? Do you want to discuss the findings? Are you a researcher and you need access to the study’s details? Reach out and we will arrange for it.